Design and delivery of sport for all programmes: should it be market, non-profit, or state-run?

Abstract

Based on a survey conducted in the city of Munich with a sample of n = 6924 residents and data on sport programmes, multilevel analyses were carried out to investigate the design and delivery of sport programmes and whether substitution effects exist. The results suggested that sport programmes offered by non-profit sport clubs foster participation in clubs, whereas the availability of programmes of commercial providers decreases participation in non-profit clubs. Thus, substitution effects were observed. However, commercial providers also complemented non-profit clubs in terms of specific sport programmes, as different groups are targeted and wider hours of operation are provided. No influence was observed for state-run programmes. Thus, non-profit sport clubs can be regarded as guarantors of sport participation, and, since they also foster social capital and positive externalities, policy makers should invest public funds in non-profit sport clubs.

Keywords:

• non-profit sport clubs
• commercial sport provider
• municipal sport programmes
• sport participation
• physical activity

Notes

1. Sport participation was defined as being physically active in one’s leisure time (including leisurely activities like casual swimming, walking the dog, but not gardening) at least once a week for 30 minutes (Downward and Rasciute 2010, Van Tuyckom and Scheerder 2010). This perspective is derived from the 1992 European Sports Charter which indicates that sport is associated not only with competition but also with casual or organised participation. Using an American perspective, similar broad definitions, even including gardening have been used, e.g. Humphreys and Ruseski (2006, 2007)).

2. For instance golf and equestrian activities are both costly and time-consuming. Thus, it can also be assumed that this scheme (expensive and time-saving activities on the one hand and cheap and time-consuming activities on the other hand) is not valid for all circumstances and sports. Preferences might interfere with these previous findings.

3. Since we have only 25 districts (25 cases at the context level) we decided not to run models with interaction effects between levels and display them in the results section. However, out of interest, we estimated a model with the DV commercial sport participation and used female, gender and working time on the individual level and created an interaction with the number of commercial sport providers per district at the context level. The effects of all four variables (incl. the two interactions) were significant. The effect for being female was positive and the interaction of being female and commercial programmes was negative. The effect of working time was positive and the interaction of working time and commercial programmes was negative. This would mean that females are more likely to use a commercial provider, however that the more programmes are offered the more likely it is that males use commercial providers. Individuals with high working time favour commercial sport providers, however, if the number of programmes increases, they are less likely to use commercial providers (maybe because the supply is too big and they cannot decide). These are interesting results, but sample size is definitely a concern. Therefore, we would like to encourage future research to include those interactions.

4. Multilevel modelling is still a rather new approach analysing sport participation and not all research that used this technique indicated the values for R² (e.g. Haug et al. 2010) which is negligent or research only presented the ICC as indication for the variance of the second level (e.g. Todd et al. 2005, van Tuyckom and Scheerder 2010).