New sports stadia for Africa? The impact of sportscape features on attendance intentions in sub-Saharan African club football

ABSTRACT

Research Question

Although team sports, particularly football, are extremely popular in sub-Saharan Africa, numerous major investments in the region’s sports infrastructure have failed, often because of sub-optimal requirements planning and low spectator demand. In light of FIFA’s recently announced aim to considerably invest into football stadia in Africa and in response to the lack of research into spectator demand in emerging countries, the present study assessed the impact facility-related factors on stadium attendance intentions in sub-Saharan Africa.

Research Method

Football fans in Ghana, Nigeria, South Africa, and Zimbabwe (N = 1,005) completed a questionnaire on facility-related factors that might affect their intentions to attend games. Structural equation modeling of the resulting data indicated which factors impact spectator behaviors.

Results and Findings: Not all the factors with the greatest impact on spectator behaviors in the four countries studied are the same as those identified in developed economies. In addition, each country’s specific characteristics appear to influence the extent to which different variables impact intentions to attend games through satisfaction with the stadium, satisfaction with the team, and trust in the team.

Implications

The study provides a better comprehension of facility-related variables and their effects on spectator demand in sub-Saharan Africa, a region which has seen very little investigation, despite the popularity of sports in this part of the world. In addition, the study is one of among a few that analyses the relevance of trust for purchase intentions in a sports setting.

KEYWORDS:

• Servicescape
• repurchase intention
• satisfaction
• trust
• sports demand

Introduction

The Fédération Internationale de Football Association (FIFA) is planning to invest a billion US dollars to promote and grow association football in Africa, a continent with a young¹ and rapidly growing population that includes a great number of football fans (FIFA, 2021; Gondwe, 2019). FIFA hopes this investment will enable local stakeholders to build stadia and other football infrastructure. However, many previous investments in sports infrastructure in Africa, especially sub-Saharan Africa, have been far from successful (Alm et al., 2016). For example, most of the new stadia built for the 2014 Football World Cup in South Africa, the first – and, so far, only – World Cup to have been held in Africa, have proved to be white elephants, as they are of no practical use to local people (Alm et al., 2016). Angola, Gabon, and Equatorial Guinea had similar experiences when they hosted or co-hosted the 2010, 2012, and 2017 editions of the Africa Cup of Nations (Dubinsky, 2021; Richardson, 2017). The reasons for these failures range from suboptimal requirements planning, politics (Alm et al., 2016), and corruption (Stathopoulou et al., 2021) to poor maintenance, insufficient funding, and – particularly – low spectator demand (Dubinsky, 2021).

FIFA’s motivation for investing in Africa’s sports facilities is partly political (Panja, 2018; Richau et al., 2019), but it also wants to improve local clubs’ finances and competitiveness by raising the quality of African football, enhancing the experience of attending local matches, and attracting more spectators to games (FIFA, 2020; Smith, 2021). A sport facility’s physical environment appears to be an important factor in encouraging return visits and in boosting sports fans’ attendance intentions (Shin & Lyu, 2019), at least in developed countries. However, the issue has never been investigated in developing countries, including those of sub-Saharan Africa, even though the factors that influence the consumption of sport events may vary greatly between industrialized and emerging economies (Buraimo et al., 2018). In addition, the poor quality of sub-Saharan Africa’s sports infrastructure (Pannenborg, 2012) may also affect the links between sportscape features and attendance intentions, so it is an ideal region in which to examine whether the same variables impact attendance intentions in emerging countries as in developed countries with modern sports infrastructure. Football is the most popular sport in Africa and therefore an obvious setting in which to investigate this issue.

The current study addressed this research gap by analyzing survey data from four sub-Saharan African countries – Ghana, Nigeria, Zimbabwe, and South Africa. Its aim was to develop an integrated framework for investigating factors that influence spectator demand and evaluating their impact. From a theoretical perspective, the study identifies the effects of several facility-related variables on satisfaction with the stadium, satisfaction with the team, trust in the team, and, ultimately, intentions to attend games. Although trust is a potentially important variable that is known to affect consumers’ repurchase intentions (Pennington et al., 2003), studies of its impact in sport settings are rare. Lastly, this analysis of sports fans’ stadium attendance intentions in four sub-Saharan countries adds to the literature by providing evidence from developing and emerging regions, as nearly all prior research on this subject has focused on industrialized countries. From a practical perspective, the study’s findings may help international organizations, local sport associations, policymakers, and sport managers to ensure that sport infrastructure projects in sub-Saharan Africa take into account both the needs of local supporters and the factors that increase stadium attendance.

Literature review and research hypotheses

Drivers of intentions to attend sport events

Economists and marketing researchers have investigated possible influences on sports attendance since the field of sports management first emerged (e.g. Demmert, 1973). Schreyer and Ansari's (2021) literature review lists more than 195 econometric analyses of the determinants of attendance, alone, and numerous other studies have explored how a wide range of sociological and psychological factors interrelate to affect spectator demand and intentions to attend sport events (e.g. Yoshida, 2017). Nevertheless, very few studies have looked at attendance intentions in developing countries.

The term ‘developing country’ covers a wide range of economic situations, ranging from ‘least developed countries’ to ‘emerging market economies’. Emerging market economies have grown rapidly and now have some (but not all) of the characteristics of developed markets. These are countries whose ‘industries have undergone and are continuing to undergo dramatic structural changes, and whose markets hold promise despite volatile and weak legal systems’ (Luo & Tung, 2007, p. 482). The term ‘frontier markets’ is used to describe a stage of economic development that lies between least developed countries and emerging market economies. Although there is no universally accepted definition of what constitutes a frontier market, the term is generally applied to countries that have begun to develop but whose economies are too small or too risky to be classified as emerging markets (Opute & Madichie, 2017). According to Speidell (2011), Africa, Asia, Europe and South America contain around 60 frontier markets.

Examples of these three types of economy in sub-Saharan Africa include Zimbabwe, a least developed country; South Africa, an emerging market economy; and Ghana and Nigeria, both of which are frontier markets. In all four cases, sport is restricted by the country’s lack of economic development and poor infrastructure (Buraimo et al., 2018).

The present study built on previous work on spectators’ demands to develop an integrative framework for exploring the premise that certain facility-related variables encourage spectators to attend sport events (Foroughi et al., 2019). As the conceptual model shows (see Figure 1), the main drivers of intentions to attend football games were expected to be trust in the team, satisfaction with the team, and satisfaction with the stadium. Factors linked to the facility (i.e. seat space, safety, cleanliness, stadium access, concessions, operating time, crowding) directly influence an individual’s satisfaction with the stadium, whereas other factors (i.e. perceived value, team identification) may influence intentions to attend games via their effects on satisfaction with and trust in the team and/or satisfaction with the stadium.

Figure 1. Conceptual model.

Effects of trust in the team

Trust, which Morgan and Hunt (1994) defined as confidence in an exchange partner’s reliability and integrity, has been recognized as one of the most important drivers of long-term relationships since these same authors developed their commitment–trust theory. For customers, trust describes their confidence in a company’s reliability and integrity, and this confidence impacts their willingness to engage with that company (De Wulf et al., 2001). Buyers’ trust in service-providing companies is a key antecedent of business transactions (Pennington et al., 2003). Trust is also a crucial factor in maintaining the relationship between customer and service in the case of high-credence services, where the outcome of the service is unknown and difficult to evaluate before consumption (Belanche et al., 2014; Eisingerich & Bell, 2007). Sport events are similar to high-credence services in that fans cannot know an event’s outcome before attending. This observation led Wu et al. (2012, p. 180) to suggest that ‘the extent to which fans trust the team or the player influences their willingness to attend the games.’ In general, research on the sports sector has connected trust with loyalty (Tsiotsou, 2013), team identification (Wu et al., 2012), and intentions to attend a sporting event (Kim et al., 2011; Magnusen et al., 2012). For example, Filo et al.'s (2008) analysis of how brand trust impacts the relationship between distinct brand associations and brand loyalty in the sports and fitness industry led them to conclude that brand loyalty will not develop until brand trust is in place. In a conceptual paper, Kim and Trail (2011) suggested that building a trusting relationship with supporters will allow a sport organization to draw more fans to its stadium and to charge higher prices.

Although the relationship between satisfaction, trust and customer intentions is well established in the marketing literature (e.g. Lu et al., 2016; Walsh et al., 2010) and has been investigated in sports-focused relationship quality literature (e.g. Lee et al., 2020; Wang et al., 2012), studies of the link between trust and intentions to attend sport events have produced contrasting results. For example, Magnusen et al. (2012) found a direct link between trust and attendance intentions, whereas Kim et al. (2011) did not find such a relationship. The present study focused on fans’ intentions to attend their favorite local football team’s games. It was hypothesized that:

Hypothesis 1: Trust in the team will have a positive effect on intentions to attend games.

Effects of satisfaction

Satisfied supporters are as important to a sport organization as satisfied customers are to companies in other industries (Van Leeuwen et al., 2002). Satisfaction can be defined as the degree to which a customer’s evaluation of a service meets their expectations (e.g. Jones & Suh, 2000) some studies have measured satisfaction with respect to a single exchange or encounter, that is, transaction-specific satisfaction, such measures do not sufficiently explain customer satisfaction, as not every encounter necessarily correlates with a customer’s overall satisfaction with a service (Bitner and Hubbert, 1994). Consequently, the present study used a measure of cumulative satisfaction, taking into account the sum of a customer’s prior experiences with a product or service (Bitner & Hubbert, 1994; Rosen & Surprenant, 1998), as such measures are likely to be better predictors of future behavior (Jones & Suh, 2000). Both, cognitive and affective components of the customers’ satisfaction phenomenon play a role in sport consumption behavior. As in the case of trust, there is a general consensus in the marketing literature that satisfaction has a positive effect on customers’ attitudes towards a service or product and ultimately determines their purchase intentions (Jones & Suh, 2000; Martin et al., 2008). Satisfaction has been linked both to stadium attendance (Bodet & Bernache-Assollant, 2011; Jones & Suh, 2000) and to consumers’ preferences in other sports marketing situations (Jang et al., 2020). In line with these findings, it was hypothesized that:

Hypothesis 2: Satisfaction with the team has a positive effect on intentions to attend games.

Hypothesis 3: Satisfaction with the stadium has a positive effect on intentions to attend games.

Numerous studies have found links between satisfaction and trust, but there is still disagreement on whether trust is a prerequisite for satisfaction (Berry, 2000) or whether satisfaction is a prerequisite for trust (Selnes, 1998). Geyskens et al. (1999) meta-analysis of satisfaction in marketing channel relationships supported this latter proposition, as their results showed satisfaction to be an antecedent to trust, with satisfaction forming in a relationship’s initial stages and trust developing in later stages. Services marketing studies have reported similar results (Han & Hyun, 2015; Leisen & Hyman, 2004), but the relationship between satisfaction and trust in sport has not been investigated empirically. In line with findings in other fields, it was hypothesized that:

Hypothesis 4: Satisfaction with the team has a positive effect on trust in the team.

Effects of facility-related antecedents

Spectators’ perceptions of a sport event’s quality depend on their assessments of the sporting performance they witness. This performance constitutes the event’s core service (Ko et al., 2011; Quansah et al., 2021), defined as a service provider’s basic competencies and the principle reason why the provider is in the market (Ferguson et al., 1999). The quality of the core service is an important determinant of consumer satisfaction, but satisfaction also depends on a consumer’s overall assessment of the relative inferiority or superiority of an organization and its services (Zeithaml, 1988). These assessments take into account facility-related factors, that is, features of the servicescape (Wakefield & Blodgett, 2016), as well as other non-core service factors and features. Thus, variables such as operating time and crowding may play an important role in determining spectator satisfaction in sport settings (Byon et al., 2013).

A servicescape is a physical environment that was designed and built to fulfill specific consumer needs (Bitner, 1992). Bitner’s (1992) typology of service organizations places modern sport events in the elaborated interpersonal services category, where the service encounter involves interaction between the service provider and customers (spectators) within the service provider’s facility (the stadium). Because the servicescape, known as the sportscape in sport settings (Wakefield & Blodgett, 1996), is the place of the service encounter and is palpable to all customers, it often has a strong cognitive, emotional, and physiological impact on customers. This impact can influence their behaviors.

Environmental psychology is the study of individuals’ responses to environmental stimuli. In the case of consumer experiences, individuals process the stimuli provided by the environment in which they consume a product or service and convert these stimuli into meaningful information, which then conditions their response (satisfaction or dissatisfaction) to the experience (McKinney, 2004). Hence, an organization must ensure that the environment it provides meets its customers’ needs and preferences.

Seating comfort includes both the comfort of seats and the space between seats. Seats may be uncomfortable because of their design (e.g. hard benches), because of their condition (wear and tear), or because they are too close together. Seating spectators too close together may impact their satisfaction with the service by making them feel physically and/or psychologically uncomfortable, cramped, and hemmed in (Wakefield & Blodgett, 1996). Yoshida and James (2010) studied the antecedents and consequences of game and service satisfaction in Japan and the United States and found that facility space (knee room, elbow room, seat arrangement, walkways and stadium space laid out so as to avoid crowding) did not significantly impact satisfaction. Conversely, Tsuji et al. (2008) found that space allocation was one of the largest contributors to spectator enjoyment at an annual state games in the southeastern United States. Hence, it was hypothesized that:

Hypothesis 5: Seating comfort has a positive effect on satisfaction with the stadium.

Stadium disasters and instances of fan violence in stadia (Cleland & Cashmore, 2016) have made stadium security an important and widely studied issue. Research highlights the importance of security features (e.g. Kelley & Turley, 2001; Yoshida & James, 2010) and suggests that safety measures in and around stadia and the presence of professional security personnel impact spectators’ perceptions of stadium security and may affect their satisfaction (Höck et al., 2010; Rocha & Fleury, 2017; Yoshida & James, 2010). Therefore, it was hypothesized that:

Hypothesis 6: Safety has a positive effect on satisfaction with the stadium.

Spectators are more likely to have a pleasant experience and to connect with their environment if it is clean (Wakefield & Blodgett, 2016). Conversely, a dirty environment can reduce spectator satisfaction (Cho et al., 2019). Indeed, as Yoshida and James (2010) first suggested and subsequent studies have shown (e.g. Hill & Green, 2012) sportscape cleanliness is closely linked to spectator satisfaction and behavioral intentions. More recently, Jang et al. (2020) analyzed links between sportscape, emotion, and behavioral intentions in North America’s major baseball (MLB), American football (NFL), ice hockey (NHL), and basketball (NBA) leagues. Cleanliness was not an important factor for basketball fans, but it impacted positive emotions and attendance intentions in the cases of baseball, American football, and ice hockey. Hence, it was hypothesized that:

Hypothesis 7: Cleanliness has a positive effect on satisfaction with the stadium.

Stadium accessibility is an important consideration for spectators wishing to attend a sport event (Wakefield & Blodgett, 1996). Most spectators are unwilling to spend large amounts of time searching for a parking space or to walk a long distance from their car to the stadium (Byon et al., 2013). Because many roads in sub-Saharan Africa are in poor condition and because rates of private vehicle ownership are relatively low (Bruun et al., 2016), stadium accessibility (whether by private vehicle or by public transportation), the state of roads to the stadium, and the availability of convenient parking are major concerns for sports fans in the region. Therefore, it was hypothesized that:

Hypothesis 8: Stadium access has a positive effect on satisfaction with the stadium.

Concessions give spectators the opportunity to buy food and drink within a stadium’s boundaries. The quality and price of this offer affect spectator satisfaction and contribute to the total fan experience (Hill & Green, 2000; Höck et al., 2010). For example, a survey of North American student spectators’ habits at college sport events showed that concessions must provide at least a moderate selection of reasonable quality food and drinks in order to satisfy stadium visitors (Trail et al., 2008). Consequently, it was hypothesized that:

Hypothesis 9: Concession quality influences satisfaction with the stadium.

Although operating time is not part of a stadium’s servicescape, it can have a significant impact on stadium attendance (e.g. Ko & Pastore, 2005). Indeed, operating time – the day of the week and time of day at which a game is played – affects how convenient it is for spectators to attend a game. For example, spectators generally have more leisure time at weekends than during the week (Buraimo & Simmons, 2015). Hence, it was hypothesized that:

Hypothesis 10: Operating time impacts satisfaction with the stadium.

Crowding is closely linked to several other sportscape variables. Being with many other people in a restricted space can generate feelings of stress that may be either positive or negative (Stokols, 1972). Despite numerous studies on the consequences of crowding, findings differ as to whether crowding positively or negatively impacts spectator satisfaction and intentions to attend games. Some studies (e.g. Machleit et al., 2000) have found that consumer crowding tends to have a negative influence on how people perceive the atmosphere in a stadium, due to crowded environments causing consumers to experience negative emotions (e.g. unhappiness, irritation) and to exhibit unfavorable purchase behaviors. On the other hand, the interplay between perceived crowding and emotions (exciting atmosphere) can generate satisfaction and promote positive purchasing behaviors (Li et al., 2009). Perceptions of crowding differ not only between individuals but also according to the environment in which the crowding occurs. Sport events, alongside clubs, bars, and festivals, are environments in which people expect and generally welcome crowding (Desor, 1972). Although some studies have found that crowding generates negative stadium experiences (Wakefield & Blodgett, 1996), it is likely that most people who attend events such as football games expect and want stadia to be crowded. In fact, the enthusiasm of a large number of sports fans is one of the most exciting and important components of a stadium’s atmosphere (Yoshida & James, 2011). Hence, it was hypothesized that:

Hypothesis 11: Crowding has a positive effect on satisfaction with the stadium.

Effects of perceived value

Consumers base their perceptions of a service or product’s value by evaluating its quality/benefits in relation to its cost (Zeithaml, 1988). In the case of sport, spectators assess the cost–benefit ratio of games they attend and of any tickets, merchandise, or food and drink they purchase (Kwon et al., 2007). Although it would seem logical for consumers to feel greater satisfaction when they perceive a service as good value for money, the link between perceived value and satisfaction has been subject to debate (Cronin et al., 2000). Nevertheless, Cronin et al. (2000) found strong evidence across several market sectors that value perception can influence satisfaction, and customers who feel they have received good value for money report greater satisfaction than those who do not (Zeithaml, 1988). Studies in a variety of industries have found that perceived value also impacts trust between supplier and customer (e.g. Yuen et al., 2018). Consequently, it was hypothesized that:

Hypothesis 12: Perceived value has a positive effect on satisfaction with the team.

Hypothesis 13: Perceived value has a positive effect on satisfaction with the stadium.

Hypothesis 14: Perceived value has a positive effect on trust in the team.

Effects of team identification

According to social identification theory (Ashforth & Mael, 1989), individuals define themselves and their social identity through their membership of a social group. Many sports fans identify very strongly with and define themselves in relation to the team they support (Bodet & Bernache-Assollant, 2011; Wann, 2006). This team identification can be considered the sporting equivalent of organizational identification. Identification simultaneously allows people to differentiate themselves from outgroup members, who do not share their characteristics, and to feel closer to ingroup members, with whom they have characteristics in common (Van Knippenberg & Van Schie, 2000). Individuals who identify very strongly with an organization are more likely to adopt the organization’s perspective and to think and act in terms of this group membership (Mael & Ashforth, 1992). Therefore, team identification can be defined as ‘the extent to which individuals perceive themselves as fans of the team, are involved with the team, are concerned with the team’s performance, and view the team as a representation of themselves’ (Branscombe & Wann, 1992, p. 1017).

Attachment to a specific team can greatly influence a spectator’s attitude (Mahony et al., 2002; Wann, 2006). As in other contexts, individuals’ evaluations of attitude objects associated with the self tend to be positive, and these positive evaluations create a cognitive- and affective-driven feeling of satisfaction (Van Knippenberg & Van Schie, 2000). Similarly, the feeling of having shared goals and objectives helps to generate trust in an organization. This effect has been demonstrated in a variety of settings, including agriculture (Ole Borgen, 2001) and public sector organizations (e.g. Hameed et al., 2013).

Supporters not only identify more strongly with their team when it wins, they also feel greater enjoyment and satisfaction (Gau et al., 2007; Madrigal, 1995). Supporters who identify strongly with a team tend to feel part of the team’s positive achievements, and this feeling can intensify their enjoyment and pleasure (Madrigal, 1995). Consequently, it was hypothesized that:

Hypothesis 15: Team identification has a positive effect on satisfaction with the team.

Hypothesis 16: Team identification has a positive effect on trust in the team.

Method

Sample and procedure

Data to test these hypotheses were collected via an online questionnaire that was made available to football fans in Nigeria, Ghana, Zimbabwe, and South Africa. Because statistics on average football match attendances in sub-Saharan Africa are unavailable, these four countries were selected on the basis of how well their clubs had performed in continental football competitions over the previous five years, as indicated by their positions in the Confederation of African Football’s (CAF) Champions League rankings. In Europe, whose continental association (UEFA) uses a similar ranking system (UEFA association club coefficient), clubs in the most successful associations (England, Germany, Italy, Spain, France) have the highest mean attendances, whereas clubs in the least successful associations (San Marino, Andorra, Montenegro) have the smallest mean attendances. CAF rankings range from 0 (low performance) to 100 (high performance), with a median value of 11.5 for Sub-Saharan African countries. South Africa and Nigeria are examples of highly successful football associations because their CAF rankings (76.5 and 32.5, respectively) were well above this median value. Zimbabwe and Ghana are examples of less successful associations because their CAF rankings (9 and 8, respectively) were below the median value.

The final sample was compiled using a total enumerative sampling method. As a first step, respondents were recruited by posting links to the questionnaire and creating subject pools on each country’s most popular football-related news websites. After removing registrations with identical email addresses from the subject pools, links to the online survey were sent through Qualtrics to a random sample created from the remaining pool members. Given the low literacy and urbanization rates in Sub-Saharan Africa (The World Bank, 2021), the rarity of online panels, and the lack of information on the number of football fans in Sub-Saharan Africa, this approach was felt to be the best way of obtaining a valid sample. Of the 3,204 people who registered to participate in the survey, a random sample of 2,000 participants was asked to complete the questionnaire. Almost two-thirds (1,299) of these participants returned completed questionnaires. However, participants who failed to answer the two items asking them to indicate their nationality and to type the name of the local team they supported were excluded from the analyses.

To minimize the number of careless respondents included in the final sample, the questionnaire incorporated several attention check questions and included a minimum time limit (Smith et al., 2016). As a final careless respondent screening check, the questionnaire followed Meade and Craig’s (2012) recommendation to include the self-report measure: ‘In your honest opinion, should we use your data?’ Eliminating respondents who failed these criteria gave a final, non-probabilistic sample of 1,005 respondents, with approximately equal numbers of respondents from each country. Almost half of the respondents had supported their team either for between 11 and 20 years (26.6%) or for between 6 and 10 years (22.1%). The remaining respondents had supported their team for 21 years or more (18.2%), for between 3 and 5 years (16.4%), or for 2 years or less (16.7%). Table 1 provides demographic data for the respondents from each country.

Table 1. Demographics.

Female	6.5%	24 years or less	30.0%
Male	93.5%	25–34 years	49.1%
		35 plus years	20.9%
Not completed secondary education	6.5%	Obtained a bachelor’s degree	36.7%
Completed secondary or tertiary education	51.3%	Master’s or doctoral degree	5.6%
Nigeria	249	Ghana	248
-earn <10k NGN per month	11.6%	-earn <500 GHS per month	30.2%
-earn between 10k and 99k NGN	59.4%	-earn between 500 and 1,999 GHS	54.0%
-earn >100k NGN	28.8%	-earn >2,000 GHS	15.6%
Zimbabwe	254	South Africa	254
-earn less than 10k ZWD per month	84.3%	-earn < 2.5k ZAR per month	26.4%
-earn between 10k and 199k ZWD	8.7%	-earn between 2.5k and 19.9k ZAR	53.9%
-earn 200k ZWD and more	7.2%	-earn 20k ZAR and more	19.8%

Measures

All the key constructs were assessed using established measures, as listed in Table 2. Participants gave their responses to all these measures, except for ‘Satisfaction with team’ and ‘Satisfaction with stadium’, on 7-point Likert scales from ‘strongly disagree’ to ‘strongly agree’. They gave their responses to ‘Satisfaction with team’ and ‘Satisfaction with stadium’ following Cronin and Taylor (1992), with bi-polar scales by Andrews and Withey (1976) and Flint et al. (2011). Responses to the cleanliness items were reverse-coded because this factor is a negative response and was expected to have a negative impact on satisfaction.

Table 2. Estimates and validity.

	Factor loadings	Cronbach’s alpha	CR	AVE
Seating comfort – adapted from Fernandes and Neves (2014) The seat arrangements at X’s stadium provide plenty of space.	.806	0.781	0.792	0.560
X’s stadium provides plenty of knee room in their seats.	.736
X’s stadium provides comfortable seats.	.699
Cleanliness (reverse coded) – adapted from Barber and Scarcelli (2010) I have chosen not to attend a game of X based on the cleanliness of its stadium.	.922	0.818	0.871	0.775
I have chosen not to return to a stadium based on the cleanliness of the stadium.	.752
Safety – adapted from Rocha and Fleury (2017) I feel safe going to X's stadium	.872	0.924	0,924	0,803
I feel safe inside X's stadium	.922
I feel safe around X's stadium	.894
Facility access – adapted from Wakefield and Blodgett (1996) X’s stadium has ample parking.	.644	0.851	0.862	0.612
There are ample public and private transport options to get to X’s team stadium.	.782
Public and private transport make it is easy to get in and out of the stadium.	.846
Public and private transport is convenient to/from X’s stadium.	.841
Concessions – adapted from Biscaia et al. (2013) There are specific foods at the stadium that I like to eat.	.675	0.858	0.869	0.693
I enjoy eating at the stadium.	.948
Eating and drinking at the stadium is something that I like to do.	.851
Operating time – adapted from Ko and Pastore (2005) The operating hours of the games are convenient.	.775	0.877	0.882	0.714
Game times are convenient.	.925
The scheduled times for watching local football games are convenient.	.829
Perceived value – adapted from Kwon et al. (2007) What I get from attending games of X is worth the cost.	.738	0.812	0.817	0.599
All things considered (price, time, and effort), buying a ticket to see X is a good buy.	.819
Compared with other leisure activities, visiting a game of X is good value for money.	.762
Crowding – adapted fromYoshida and James (2011) Being surrounded by thousands of fans at a game is a great experience.	.750	0.737	0.722	0.566
I am excited by being with other fans who are cheering, yelling, singing, and screaming for their team.	.754
Team identification – adapted from Mael and Ashforth’s (1992) When someone praises X, it feels like a personal compliment.	.703	0.833	0.838	0.510
X’s success is my own success.	.663
When I talk about X, I usually say ‘we’ rather than ‘they.’	.726
When someone criticizes X, it feels like a personal insult.	.815
If a story in the media criticized X, I would feel embarrassed.	.650
Team performance – adapted from Yoshida and James (2011) X’s players perform well-executed plays.	.775	0.837	0.839	0.512
The players of X have superior skills.	.663
X gives 100% in every game.	.712
X plays hard all the time (put lots of effort).	.737
The players of X always try to do their best.	.685
Satisfaction with the team – adapted from Cronin and Taylor (1992) Overall, my feelings towards X’s performance can best be described as … (unsatisfied–satisfied).	.907	0.952	0.953	0.837
Overall, my feelings towards X’s performance can best be described as … (displeased–pleased).	.939
Overall, my feelings towards X’s performance can best be described as … (unhappy–happy).	.930
Overall, my feelings towards X’s performance can best be described as … (terrible–delighted).	.882
Satisfaction with the team’s stadium – adapted from Cronin and Taylor (1992) Overall, my feelings towards X’s stadium can best be described as … (unsatisfied–satisfied).	.922	0.966	0.967	0.880
Overall, my feelings towards X’s stadium can best be described as … (displeased–pleased).	.951
Overall, my feelings towards X’s stadium can best be described as … (unhappy–happy).	.956
Overall, my feelings towards X’s stadium can best be described as … (terrible–delighted).	.922
Trust in the team – adapted from Walsh and Beatty (2007) My team can generally be trusted	.780	0.929	0.923	0.667
I trust my team	.844
My team has high integrity/	.837
I can depend on my team to do the right thing every time they play.	.857
My team can be relied upon giving the best in every game.	.803
I have great confidence in my team.	.775
Intention to attend – adapted from Jones et al. (2000) How likely is it that you will attend a football game of X at their stadium again in the future?	.828	.902	.904	.759
How probable is it that you will attend a football game of X at their stadium again in the future?	.901
How possible is it that you will attend a football game of X at their stadium again in the future?	.883

Because spectator behaviors and attitudes such as match day attendance, satisfaction, and trust in the team are closely related to spectators’ overall evaluations of the game experience through team performance (Biscaia et al., 2013; Yoshida & Nakazawa, 2016), team performance was used as a control variable to circumvent potentially spurious findings and thus establish more accurate estimates of the paths between the variables in the hypothesized model (Becker et al., 2016).

Results

Measurement model, common method variance, and measurement invariance

The measurement model was tested by conducting a confirmatory factor analysis (CFA) combined with the robust maximum likelihood (MLR) estimation method. Elevated inter-correlated error terms showed that some items on a scale were redundant and that it could be simplified to improve the fit. After these modifications, the model showed good fit: (χ2(1130) = 2644.432, p < .001; comparative fit index [CFI] = .961; Tucker–Lewis Index [TLI] = .956; standardized root mean square residual [SRMR] = .0387; and root mean square error of approximation [RMSEA] = .035). Table 2 shows reliability and validity statistics for the measures used. All the measures had good reliability, as Cronbach’s alpha and composite reliability (CR) coefficients were higher than .07 and the average variance extracted (AVE) was greater than .05 (Bagozzi & Yi, 1988).

Common method variance (CMV) occurs when one factor accounts for most of the covariance among a set of variables. A Harman’s single-factor test in which all the model items were entered into an exploratory factor analysis with principal axis factoring (unrotated) showed that this combined factor explained 25.814% of the variance. This is below the 50% threshold, thereby indicating that CMV is not a concern (Podsakoff et al., 2003). As a further check of whether the variables were affected by CMV, a series of CFA marker models were tested, as recommended by Williams et al. (2010). For these tests, color scale – a scale unrelated to the rest of the model’s substantive variables – was used as an ideal marker (Simmering et al., 2015). The baseline CFA model included the 15 substantive variables as well as the marker variable. Correlations between the substantive variables and the marker variable were set to 0 and the regression weights and variances of the marker variable were fixed to the values obtained from the CFA marker model. The constrained model (Model C) was built by prefixing the factor loadings from the marker variable to all the substantive variable indicators, which were set to be equal. The next step was to build an unconstrained model (Model U) in which the factor loadings from the marker variable to all the indicators were unconstrained and freely estimated. Finally, the restricted model (Model R) was built by fixing the substantive factor correlations to values from the baseline model. Model C did not give a significantly better fit than the baseline model, indicating the absence of CMV. Table 3 provides a summary of the model results.

Table 3. CMV – model comparison.

Model	χ^2	Df	TLI	CFI	RMSEA
Baseline	2689.196	1288	.958	.963	.033
Method C	2687.542	1287	.958	.963	.033
Method U	2608.374	1237	.958	.963	.033
Method R	2687.582	1378	.964	.965	.031
	Chi-Square Model Comparison Tests
ΔModels	Δχ^2	Δdf	p-value
Baseline vs. Method C	1.654	1	.198415
Method C vs. Method U	79.168	51	.006956

Hypotheses testing

Structural equation modelling (SEM) was used to test the relationships between the data and the conceptual model, using the maximum likelihood estimation method in SPSS Amos 27.

The SEM model fit the data well (χ2(1150) = 2589.087, p < 0.001; RMSEA = 0.035; CFI = 0.960; TLI = 0.956; SRMR = 0.0600). When looking at the four countries individually, the fits between the data and the model were: Ghana: χ2(1150) = 1946.789, p < 0.001; RMSEA = 0.053; CFI = 0.903; TLI = 0.892; SRMR = 0.0693; Nigeria: χ2(1255) = 1912.066, p < 0.001; RMSEA = 0.052; CFI = 0.921; TLI = 0.913; SRMR = 0.0752; Zimbabwe: χ2(1150) = 1912.653, p < 0.001; RMSEA = 0.051; CFI = 0.922; TLI = 0.913; SRMR = 0.0776; and South Africa: χ2(1255) = 1900.333, p < 0.001; RMSEA = 0.051; CFI = 0.925; TLI = 0.917; SRMR = 0.0715.

Table 4 shows results of the SEM analysis for the hypothesized relationships for the whole sample and for each of the four countries.

Table 4. Results of SEM.

Relationships	Whole sample-β	Zimbabwe	Ghana	Nigeria	South Africa
(H1) Trust in team → Intention to attend game	.335***	.341***	.170+	.207*	.495***
(H2) Satisfaction with stadium →Intention to attend game	.179***	.184*	.173*	.226**	.232***
(H3) Satisfaction with team → Intention to attend game	.029	.083	.232*	.152	.194*
(H4) Satisfaction with team → Trust in team	.228***	.278***	.270***	.183***	.207**
(H5) Seating comfort →Satisfaction with stadium	.464***	.467***	.352***	.220+	.310*
(H6) Safety → Satisfaction with stadium	.064	.061	.063	.162	.184
(H7) Cleanliness → Satisfaction with stadium	.148***	.104	.234***	.157+	.052
(H8) Facility access → Satisfaction with stadium	.009	.057	.109	.011	.046
(H9) Concessions → Satisfaction with stadium	.137***	.111+	.093	.175*	.064
(H10) Operating time → Satisfaction with stadium	-.035	-.024	-.027	.013	-.161
(H11) Crowd experience → Satisfaction with stadium	-.111**	-.019	-.237**	-.089	-.112
(H12) Perceived value → Satisfaction with team	.087	-.042	.408**	.189	-.109
(H13) Perceived value → Satisfaction with stadium	.081	.039	.325**	.059	.056
(H14) Perceived value → Trust in team	.225***	.254**	.163+	.239**	.271*
(H15) Team identification → Satisfaction with team	.121***	.207**	.031	.420***	-.006
(H16) Team identification → Trust in team	.337***	.267***	.233**	.453***	.386***
Team performance → Satisfaction with team	.453***	.577***	.318***	-.022	.752***
Team performance → Satisfaction with stadium	.091**	.129	.008	.093	.129
Team performance → Trust in team	.146***	.137	.288***	.002	.096
Team performance → Intention to attend game	-.005	-.195*	.162+	.043	.044
Total effects
Seating comfort → Intention to attend game	.083**	.087*	.061*	.05+	.071*
Safety → Intention to attend game	.012	.011	.011	.037	.034
Cleanliness → Intention to attend game	.026**	.019*	.041**	.036*	.012
Facility access → Intention to attend game	.002	.011	.019	.002	.011
Concessions → Intention to attend game	.024**	.021+	.016	.040*	.015
Operating time → Intention to attend game	-.006	-.004	-.005	.003	-.037
Crowd experience → Intention to attend game	-.02**	-.004	-.041*	-.020	-.025
Perceived value → Intention to attend game	.099**	.088	.205	.101	.153*
Team identification → Intention to attend game	.125**	.13**	.046	.175**	.188*

*** p < .001; ** p < .01; * p < .05; + p < .10.

Of the hypothesized exogenous variables, seating comfort (β = .464, p < .001), cleanliness (β = .148, p < .001), and concessions (β = .137, p < .001) positively influenced spectators’ satisfaction with the stadium. In contrast, crowding (β = -.111, p < .01) negatively influenced satisfaction with the stadium. These results for the whole sample support H5, H7, and H9, but do not support H6, H8, H10, and H11.

Again for the whole sample, perceived value had a significant impact on trust in the team (β = .225, p < .001), but not on satisfaction with the team (β = .087, p > .10) or satisfaction with the stadium (β = .081, p > .10). Team identification significantly influenced satisfaction with the team (β = .121, p < .001) and trust in the team (β = .337, p < .001). These results support H14, H15, and H16, but do not support H12 and H13.

All these relationships were moderated by country effects, either in terms of perimeter value or significance. In the case of Zimbabwe, seating comfort (β = .467, p < .001), and concessions (β = .111, p < .10) significantly influenced satisfaction with the stadium, whereas perceived value influenced trust in the team (β = .254, p < .01). In addition, team identification positively impacted both satisfaction with the team (β = .207, p < .01) and trust in the team (β = .267, p < .001). In Ghana, seating comfort (β = .352, p < .001), cleanliness (β  = .234, p < .001), and crowding (β = -.237, p < .01) significantly influenced satisfaction with the stadium, whereas perceived value significantly impacted satisfaction with the team (β = .408, p < .01), satisfaction with the stadium (β = .325, p < .01), and trust in the team (β = .163, p < .10). For Nigeria, seating comfort (β = .220, p < .10), cleanliness (β = .157, p < .10), and concessions (β = .175, p < .05) impacted satisfaction with the stadium, but perceived value influenced only trust in the team (β = .239, p < .01). Team identification influenced both satisfaction with the team (β = .420, p < .001) and trust in the team (β = .453, p < .001). In South Africa, seating comfort (β = .310, p < .05) significantly influenced satisfaction with the stadium, and perceived value (β = .271, p < .05) and team identification (β = .386, p < .001) influenced trust in the team.

Results for all four countries supported H4, as satisfaction with the team positively affected trust in the team (β = .228, p < .001). However, satisfaction with the team did not increase overall intentions to attend games (H3) (β = .029, p > .10), although the relationship between these two variables was significant in Ghana (β = .232, p < .05) and in South Africa (β = .194, p < .05). Satisfaction with the stadium directly and significantly affected intentions to attend games (H2), both for the overall sample (β = .179, p < .001) and for all four countries individually: South Africa: β = .232, p < .001; Nigeria: β = .226, p < .01; Ghana: β = .173, p < .05; and Zimbabwe: β = .184, p < .05. Finally, as hypothesized (H1), greater trust in the team was linked to greater intentions to attend games (β = .335, p < .001), but this effect was stronger in South Africa (β = .495, p < .001) and in Zimbabwe (β = .341, p < .001) than it was in Nigeria (β = .207, p < .05) and Ghana (β = .170, p < .10).

In terms of the impacts of the exogenous variables on intentions to attend games, seating comfort had a significant impact overall and in all four individual countries, (β_overall = .083**, β_Zimbabwe = .087*, β_Ghana = .061*, β_Nigeria = .05+, β_{South Africa} = .071*). Cleanliness had a significant impact overall and in Zimbabwe, Ghana, and Nigeria (β_overall = .026**, β_Zimbabwe = .019*, β_Ghana = .041**, β_Nigeria = .036*), but not in South Africa (β_{South Africa} = .012). Team identification had a significant impact overall and in Zimbabwe, Ghana, and South Africa (β_overall = .125**, β_Zimbabwe = .13**, β_Nigeria = .175**, β_{South Africa} = .188*), but not in Ghana (β_Ghana = .046). The total effects of the remaining exogenous variables on intentions to attend games were mixed. Overall, concessions (β_overall = .024**), crowding (β_overall = -.02**), and perceived value (β_overall = .099**) had a significant effect on intentions to attend games, but the effect of concessions was significant only in Zimbabwe and Nigeria (β_{Zimbabwe_concessions} = .021+; β_{Nigeria_concessions} = .04*), the effect of crowding was significant only in Ghana (β_{Ghana_crowding} = -.041*), and the effect of perceived value was significant only in South Africa (β_{South Africa_perceived value} = .153*). Team performance – the control variable in the model – impacted satisfaction with the team, satisfaction with the stadium, and trust in the team.

Discussion and conclusion

The current study investigated the relationships between a variety of variables and sub-Saharan African football fans’ intentions to attend games. Several facility-related variables, including seating comfort, cleanliness, concessions, and team identification, influenced spectators’ intentions to attend games in all four countries, but the sizes of these variables’ effects varied between countries. Results also showed that country specific factors moderated trust in the team and satisfaction with the team and with the stadium, and that other variables influenced intentions to attend games in some countries (i.e. crowding in Ghana, perceived value in South Africa).

These findings provide insights into how spectator satisfaction and trust form and how satisfaction and trust impact intentions to attend games. Although such relationships are well documented in the services marketing literature (e.g. Leisen & Hyman, 2004), the current study is the first detailed investigation of this issue in a sports setting. In the light of this closer examination of the satisfaction – trust – purchase intentions triangle, the main driver of stadium-ticket purchase intentions in sport would appear to be spectator trust, not satisfaction (with the team), as suggested by previous sports marketing research. In other words, fans’ trust in their team is a key factor in maintaining their relationships with their favorite clubs and has a significant impact on their enthusiasm for attending games. Furthermore, satisfaction with the stadium seems to be more important than satisfaction with the team in stimulating purchase intentions.

Seating comfort appears to be the most important factor in determining satisfaction with the stadium. For supporters, seating comfort means that stadium seats are large enough, comfortable, and far enough apart to provide sufficient knee room. Stadium cleanliness also greatly impacted spectators’ intentions to attend games. Finally, concessions affected satisfaction and indirectly influenced trust and intentions to attend games. Indeed, supporters enjoy eating during games and associate stadium visits with specific foods, so they want stadia to have concessions that provide these foods.

Other facility-related factors, including stadium access (i.e. parking, private and public transport to the stadium), safety, and operating time, did not significantly impact satisfaction with the stadium or behavioral intentions. Mean scores for stadium access and operating time were relatively high (i.e. above 5 on the 7-point Likert scales) and had small standard deviations. This high ceiling effect and low variability may explain why the impacts of these variables were statistically insignificant. These results are line with previous findings (e.g. Byon et al., 2013).

Crowding had a significant overall effect on satisfaction with the stadium and intentions to attend games. However, this link was significant for only one of the four countries, Ghana, and it was negative, not positive as hypothesized, possibly due to crowding generating negative feelings by hindering spectators’ ability ‘to explore and encounter stimulating experiences’ (Wakefield & Blodgett, 1996, p. 69). The negative link between crowding and satisfaction may have been stronger in Ghana than in the other three countries because of a vividly remembered sporting disaster: In 2001, 127 people lost their lives in Accra Sports Stadium when police fired tear gas during a match between Ghana’s two most popular clubs (Bowley et al., 2004). Research in environmental psychology shows that culture also shapes individuals’ reactions to crowding (Pons & Laroche, 2007).

In line with previous research, perceived value influenced trust (e.g. Yuen et al., 2018), but it did not impact satisfaction with the team (e.g. Kwon et al., 2007), except in Ghana, where perceived value had a greater impact on satisfaction with the team and on satisfaction with the stadium than it had on trust in the team. Spectators who felt that ticket prices offer value for money had greater trust in their team and were more likely to attend games. Team identification also had a large impact on spectator satisfaction, trust in the team, and intentions to attend games.

Theoretical contribution and managerial implications

The present study’s main contribution is to extend research on customer behavior to four sub-Saharan countries. Despite the abundance of sports management and sports marketing research, few studies have investigated spectator demand in developing countries. The present study helps fill this gap by showing that trust in the team and satisfaction with the stadium are important drivers of sport event attendance and that local characteristics moderate the effects of these variables.

From a managerial perspective, the study’s results provide a better understanding of the effects of facility-related variables on spectator demand. Although some of the factors that impact intentions to attend games in developed countries also apply to the countries in this study, other factors do not. Consequently, international sport federations, national federations, and clubs should not base plans for stadium and game infrastructure in sub-Saharan Africa solely on Europe’s experiences.

The four countries analyzed in this study are characterized by some economic differences. Frontier market South Africa, which hosted the FIFA world-cup in 2010, built or renovated 25 contemporary stadia ahead of the event (Death, 2011; Molloy & Chetty, 2015), while Zimbabwe, a least-developed country with per capita GDP of USD1360², has been without a modernized stadium eligible to carry out international games since 2000 (Kurira, 2021). In-between these two extremes are the emerging market countries Ghana and Nigeria, which possess 4 (Ghana) and more than 10 (Nigeria) more recent stadia (World of Stadiums, 2022) – in addition to several subpar quality ones, in which local top-tier league matches are regularly played (Pannenborg, 2012).

An interesting question is whether spectator behavior in the countries that find themselves in similar economic development stages, and are geographically close to each other – Ghana and Nigeria – showcase more similarities on how facility-related features influence soccer game attendance than the other two countries. Looking at the results, this is rather not the case. On the contrary, Ghana differs from the other three countries in that Cleanliness and Crowd experience are important factors, while perceived value drives satisfaction with team and stadium. In the other countries, perceived value rather drives trust in team. Also, team performance appears to be more important than in the other countries.

It is important that managers look at the idiosyncratic circumstances of individual countries and analyze the attendance behavior of local spectators before new stadia are being build. Solely looking at economic similarities between the countries in a region might lead to wrong conclusions. There are, however, some common denominators among the countries in the study. FIFA, local sport authorities, and sport managers should pay greater attention to seating comfort and concession stands when planning/running sports facilities in sub-Saharan Africa, and they must ensure stadia are kept clean, so spectators are not deterred from attending games.

Furthermore, not all fans appreciate crowding. Managers need to be aware that to increase spectator demand, measures need to be in place that guarantee safety and comfort to relieve spectators from the uneasiness of being among huge crowds. Some solutions could be similar to the consequences drawn by the English government following the Taylor Report in 1990. They could include replacing old manual iron turnstiles by modern entrance systems, stricter entrance – as well as alcohol controls, fixed and numbered seats, venue operations center with a good stadium overview, the introduction of family zones, and electronic ticketing systems, as well as proper police – and security training. These measures, once implemented, need to be prominently promoted in public.

Club executives need to consider both spectators’ demands and local particularities when drawing up strategies and when targeting their offers at different market segments. Sport managers can strengthen fans’ identification with their team by offering services at a fair price and, at the same time, by ensuring their facilities provide comfortable seating, sufficient concession stands offering good-quality refreshments, and adequate protection from overcrowding. As well as increasing fan satisfaction, clubs should also focus on building trust. Satisfaction with the stadium and trust in the team are the most important factors for encouraging fans to attend games. Lastly, stadium maintenance has been a large issue in sub-Saharan Africa, overcoming this issue must be a priority when pursuing investments such as those envisaged by FIFA.

Limitations and further research

Because country-by-country attendance figures for football games in sub-Saharan Africa are unavailable, the countries included in the present study were selected on the basis of the CAF’s Champions League rankings, which can be considered an indicator of demand. Although the four countries surveyed include some of the region’s most populated countries and largest economies, they do not necessarily represent the vast ethnic, cultural, and historical range covered by sub-Saharan Africa’s almost 50 countries and 900 million inhabitants. Thus, the empirical findings presented here may reflect Nigeria, Ghana, South Africa, and Zimbabwe’s specific environmental factors and characteristics and may not be representative of the situation throughout sub-Saharan Africa.

The current study focused on football because it is the most popular sport in Africa, but the relationships between constructs may differ between sports. Consequently, it would be interesting to conduct similar studies in sports that are just beginning to penetrate the region, such as basketball, as these sports also need adequate facilities. Finally, the current study examined attitudes, perceptions, and intentions, but it did not measure participants’ actual behaviors. Because behavioral intentions do not always translate into actual behaviors, future studies should also measure whether participants actually attend games.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 40 percent of Africa’s population is below 15 years of age.

2 2022 GDP per capita of South Africa was USD5900, of Ghana USD2080, and of Nigeria USD2420 Tradingeconomics. (2022). GDP per country 2021. Retrieved 23.07.2022 from https://tradingeconomics.com/country-list/gdp