Unexpected losses of local teams lead to a small decrease in the number of births nine months thereafter due to reduced short-term sexual interest and intercourse

Soccer Scores, Short-Term Mood and Fertility. Fabrizio Bernardi & Marco Cozzani. European Journal of Population, Apr 14 2021, https://rd.springer.com/article/10.1007/s10680-021-09576-2

Abstract: Previous research has shown that seemingly irrelevant events such as unexpected outcomes in sporting events can affect mood and have relevant consequences for episodes of crime and violence, investing behavior and political preferences. In this article, we test whether mood shocks associated with unexpected results in soccer matches in Spain affect fertility. We use data on betting odds and actual scores to define mood shocks and link them to births by month and province in Spain, between 2001 and 2015. We find that unexpected losses of local teams lead to a small decrease in the number of births nine months thereafter. The effect is larger for more unexpected losses, in those provinces with the largest amount of support for the local team and robust to a number of placebo tests. We argue that these results are consistent with the gain–loss asymmetry predicted by prospect theory.

Conclusions

Previous studies have documented that seemingly irrelevant events may have important consequences for political preferences and opinions, for risk-taking economic decisions and for episodes of crime and violence (Card and Dahl 2011; Edmans et al. 2007; Healy et al. 2010; Munyo and Rossi 2013). These findings have been interpreted as evidence that changes in mood spread to otherwise unrelated dimensions such as evaluation of politics or of economic risk and can trigger other types of behaviors. In this article, we build on this literature and test the hypothesis that mood shocks might influence fertility behavior. To this end, we analyze the universe of births data in Spain between 2001 and 2015 and focus on mood shocks arising from soccer scores in Spain. We compare betting odds and actual outcomes of soccer games in Spain to identify exogenous mood shocks around expected outcomes.

Two previous academic articles on the effect of sport events on fertility have produced contradictory findings. The anecdotal claim of an “Iniesta generation” following the last-minute goal by the Barcelona midfielder in the UEFA Champions league semifinal against Chelsea is confirmed by Montesinos et al. (2013), while no evidence of “Super Bowl Babies” is found by Hayward and Rybińska (2017). What these two studies have in common is that they both focus on the supposed positive effect of success in a major sport event on fertility. In our study, we enlarge the explicative framework to also consider the consequence of losses. We find that an unexpected loss by the most popular soccer team in a Spanish province leads to a reduction of 0.8% in the number of births nine months later in that province. We do not find an opposite effect for unexpected wins. This finding is consistent with an asymmetric hypothesis drawn from prospect theory, stating that mood changes arise due to deviations from expected outcomes, with losses having larger effect than wins. A possible way to reconcile our findings and those by Hayward and Rybińska (2017) and Montesinos et al. (2013) is that a sport victory has to come as really unexpected with an unique collective celebration to produce an increase in the number of births, as it might have been the case for the agonic victory of FC Barcelona against Chelsea, associated with the Iniesta generation, and less so for the Super Bowl games whose outcomes tend to be more equalized a priori.

From a quantitative point of view, the point estimate of our main finding is very small. For instance, the 0.8 percent reduction in the number of births in a given province associated with one unexpected loss of the local soccer in team nine months earlier that we have documented corresponds on average to a reduction of about 49 births for each unexpected loss in a given month for the province of Madrid. The estimated effect, therefore, does not entail any consequences for the aggregate fertility rate in Spanish provinces. The decrease in the number of births nine months after an unexpected loss by the local team is likely to be compensated in the following months, by those couples who were planning to have a child. Even small-sized effect can, however, entail theoretical relevance (Elliott and Granger 2004; Bernardi et al. 2017). First, our key finding supports the idea that emotions and mood can be important determinants for fertility. Scholars should then consider how to include emotions into the increasingly popular models of planned behavior to study fertility (Ajzen and Klobas 2013; Mencarini et al. 2015). Work in close-by disciplines can provide some fruitful interdisciplinary inspiration in that direction (Elster, 1998; Massey, 2002). Second, our main finding also provides support for the prediction of prospect theory beyond its most common applications in finance, insurance and consumption-saving decisions (Barberis, 2013).

Methodologically, our study adds to a body of studies that have investigated the effect of subjective well-being on fertility. Moods and emotions are an important component of subjective well-being (Diener et al. 1999). There is now some evidence that happier people are more likely to have children and conversely that stress and poor mood might cause infertility (Aassve et al. 2012; Cetre et al. 2016; Greil 1997; Le Moglie et al. 2015; Parr 2010). Although our results refer only to short-lived mood shock, they provide critical evidence that supports a causal interpretation of the association previously found between happiness and fertility.

A major limitation of our study that makes us interpret these suggestive results with caution is that we cannot observe the intervening mechanisms between soccer scores and mood shock and between the latter and reduction in number of births. In a direct extension of this work, one could measure mood shocks with a sentiment analysis using Twitter data on province base (Mencarini et al. 2019). One could also focus on the intervening mechanism between mood shocks and fertility, i.e., reduction in sexual desire and intercourse. One could then study the effect of mood shock on some proxies for sexual arousal and intercourse, such as the internet access to porn sites (Markey and Markey 2010) or consumption of condoms and morning-after pills.

Still, these additional analyses with different indicators for mood shocks and proxies for sexual intercourse at the province level would still suffer from a major limitation that we also face in this current work, namely that we use macro-level data to test a micro-level mechanism. In this respect, future research could focus on physiological mechanisms (Bernhardt et al. 1998; van der Meij el al. 2012) and test whether testosterone change following vicarious experience of unexpected wins and losses is indeed asymmetric, so that a hormone change after unexpected losses is larger than the increase after unexpected wins. One could also look at variations in sexual interest and behaviors (Bancroft et al. 2003; Janssen et al. 2013) and analyze whether and how mood shocks related to soccer outcomes (or any other event that might affect mood) affect sexual interest and intercourse.