There is no such thing as a free run. The determining factors of sports expenditure

The current doctoral thesis focusses on the determining factors of sports participants’ expenditure. From an economic point of view this is an important research subject, as the sports industry directly contributes to Western economic welfare and accounts for an increasing share of total employment and GDP. This is not surprising, as figures indicate 64.3% of the Flemish population practices sports at least once a year. The majority of this large number of sports participants spends money on certain sports goods and services, certainly when they practice sports in a persistent way. If both the participation and the expenditure figures are taken into account, it is found that an average Flemish household with school-aged children spends €1525 on sports, while individual sports expenditure is estimated at €352 (or €548 per sports participant). Moreover, sports participation has also an important instrumental value, as it augments health, psychological and social well-being, and thus reduces healthcare costs. The results of the current research support the government in determining the factors that prevent people from consuming sports participation (Downward, Dawson & Dejonghe, 120).

In contrast to most previous research, the studies in this thesis investigate sports expenditure on a more detailed level (separate sports activities, separate expenditure categories, often-neglected background characteristics, etc.), by using sports-specific questionnaires and innovative data-gathering methodology (i.e. observation and diary approach). In the next paragraphs a short overview is given of the answers on the three main research questions that are the guiding theme in the seven papers (Chapters 4-10) of this doctoral thesis. These chapters are preceded by a general introduction to set out the conceptualisation, theoretical framework and the methodology and datasets that are used (Chapters 1-3). Finally, Chapter 11 wraps up the findings and provides implications for the government and the other players of the sports industry.

The first research question focusses on the factors that influence sports expenditure, and is a common theme in the different papers (RQ1). For the socioeconomic variables, it is unambiguously demonstrated that income positively influences expenditure on sports, while for the other variables of this group the relationship is less straightforward. Education only determines the probability that money is spent on sports, while no effect (or even a negative one) is found on the amount that is spent. Also for time scarcity only limited evidence is found, although there is a tendency towards substitution with non-sport leisure activities. Although the investigated sociodemographic variables play their role in determining sports expenditure, their influence diminishes once people decided to take part in (a specific) sports activity. For example, average expenditure by men is higher than is the case for women, but this effect tends to disappear once male sports participants are compared to female.

When specific sports activities are investigated, it is clear that the psychographic and sports-related variables determine sports expenditure to a bigger extent than the sociodemographic and socioeconomic (except for income) variables. Indeed, the way that sports participants (un)consciously express their sports identity is of particular importance for the amount of money that they spend on sports. It is for example demonstrated that performance-based attitudes, interests, motivations and behaviours have a positive effect on sports expenditure, while a negative relationship is found for people who take part because of health or social reasons.

In Chapters 6 and 7, sports expenditure is investigated from a non-aggregated point of view, resulting in significant differences between the determinants of sports activities on the one hand, and expenditure categories on the other hand. Chapters 8, 9 and 10 build further on these findings by focussing on the two most practiced sports activities in Flanders, cycling and running. This focus allows for questioning the expenditure categories in a more specific way, and to specify the background characteristics.

Because of the determining role of income in spending money on sports, the magnitude of this relationship is investigated in the second research question (RQ2). Chapters 5, 6 and 7 calculate income elasticities that stipulate the percentage change in expenditure, in response to a one percent change in income. For economic agents who are low on income, the effect of a one percent rise in income on the probability of spending money on sports is bigger, than is the case for high-income individuals. Put differently, monetary stimuli have far more effect on convincing lower income people to consume sports, than is the case for higher incomes. Nevertheless, it is the other way around for the amount that is spent, as the effect of an income-rise has an increasingly positive effect on the income elasticity.

Also for the income elasticities, relatively large differences exist between sports activities on the one hand, and expenditure categories on the other hand. For the former, the elasticities are higher for expensive (e.g. tennis, winter sports) and time-efficient (e.g. running) sports compared to other sports (e.g. walking, martial arts, cycling, fitness). For the latter, the ‘mandatory’ (e.g. footwear, equipment, clothing) and ‘core’ (sports events, membership fees) sports goods and services tend to have higher elasticity values than the non-necessary ‘indirect’ expenditure categories (e.g. sports holiday, transport by car, sports food and drinks, additional sports lessons).

Although the overall (Chapters 4 and 5) and non-aggregated (Chapters 6, 7 and 8) survey-based studies provide in a detailed insight in the sports expenditure patterns, questionnaires have certain drawbacks such as recall bias, tactical answers, obtrusiveness, non-response, etc. Therefore, the third research question and the last two papers (Chapters 9 and 10) in this doctoral thesis explore the potential of the observation and diary methods, two methods that have seldom been applied in socioeconomic research (RQ3). First, it is shown that the observation of running event participants (Chapter 9) produces a bunch of ‘objective’ data about running apparel usage, without interference of the research subjects (and thus no non-response). As the expectations are that in the near future automatic picture analysation will be possible, these results open up interesting opportunities. Nevertheless, this also raises privacy issues, certainly because the majority of the mass sports participation events already takes pictures of its participants. Second, Chapter 10 demonstrates that diary data allow for analysing sports apparel usage at a very non-aggregated level. It is for example shown that runners wear a more expensive outfit when they take part in a running event, compared to other running sessions.

The different chapters in the current doctoral thesis demonstrate that the governmental subsidizing policy contributes to lowering the income barrier for sports participation. Yet, the results also suggest that the subsidizing policy can be organised in a more efficient way. First, the income elasticities are higher for people who are low on income. Therefore, a budget neutral policy action that augments subsidies for these low-income agents and lowers the subsidies for high-income agents, would result in a higher average participation rate. Second, the results demonstrate that focussing solely on subsidizing sports club membership and sports infrastructure is probably not the most efficient strategy in raising sports participation rates. This is because expenses on sports club membership and admission fees only make up for a small part of total sports participation expenditure. It could therefore be effective to directly donate ‘sports-vouchers’, that can only be spent on sports-related services, to people who are low on income. These people can then decide to spend them on sports club membership, or on other sports expenses. Finally, government could also consider to differentiate between sports activities. As a low income elasticity is found for fitness (and for other sports such as walking, martial arts), government should not only apply price-reducing strategies, but also consider to focus on these sports activities in sports participation campaigns.

Despite the negative connotation that is sometimes associated with the commercial sector, it contributed significantly in raising the sports participation rates. Enterprises can use the regression results of the current study to segment sports consumers, such that they are reached more efficiently. It is for example essential that sports enterprises understand that the socioeconomic and sociodemographic variables have a different impact on the decision to spend money on sports participation, and on the amount that is spent. This implies that extrapolating their non-representative client database to the whole population could result in counterproductive marketing strategies. The sociodemographic and the socioeconomic variables are found to primarily intervene in the decision to spend money on sports or not. Nevertheless, once people took the decision to take part in sports, it is mostly the psychographic and sports-related variables (and income) that determine the amount that is spent on sports.