Chapter 4. Does ethnic heterogeneity of clubs affect member dropout? 89 4.4 Results Study sample A description of the study sample split by background for all nine playing seasons combined is presented in table 4.1. Some characteristics of the study sample are worth noting. Firstly, it should not come as surprise that the amateur football member population is relatively young and, despite an influx of (young) women into the sport (De Kwaasteniet, 2019), is dominated by males. Furthermore, members with a migrant background are underrepresented in amateur football. In the study sample, 15% of the membership records belongs to members with a migrant background, while citizens with a migrant background make up roughly 21% of the Dutch population (see chapter 2)15. The numerical disparity between members with a Dutch and migrant background also translates to the club level. The ingroup share figures indicate that members with a Dutch background share their background with most other members within their club (M = 87.33, SD = 10.52), while members with other backgrounds usually share this background with a much smaller portion of the members (M = 9.30, SD = 17.71). Lastly, we also see substantially higher dropout rates for members with migrant backgrounds (23%) than for members with Dutch backgrounds (13%). In the next section, we delve further into the explanation for this difference. Hypotheses testing The outcomes of the analyses to test the hypotheses of this study are presented in table 4.2. To simplify things, the predictor coefficients from table 4.2 are also exponentiated and presented as percent probabilities in table 4.3. Model 1 includes a random intercept, controls for time and time period and all fixed individual level predictors. The results of this model show that when controlled for individual level characteristics, members with migrant backgrounds show a substantially higher probability of dropping out (i.e., 37.26%) than members with a Dutch background. In model 2, two club level predictors are added to the model: club size and club income share. While we can see that larger clubs and a higher proportion of low-income members in a club increase member’s probability to 15 It must be noted that this does not hold true for each individual background. Chapter 2 shows that some minority groups are in fact overrepresented in Dutch amateur football.
RkJQdWJsaXNoZXIy MjY0ODMw