Models of spatial voting and mobilization theory often imply that a larger number of parties should increase electoral turnout. Even though this hypothesis is well established in the political behaviour subfield, the most-up-to-date evidence arguably fails at identifying a causal effect. Employing time-series cross-sectional data, some studies show that electoral participation is higher when increasing the number of candidatures while others justify the opposite, i.e., that higher participation rates foster new party entry instead. To solve this endogeneity puzzle, I leverage a unique real-world setting that approximate the experimental ideal: the 2015 Spanish local elections. In those elections, the two newcomers Podemos and Ciudadanos run candidatures for the first time. However, they could not compete in many municipalities due to their lack of organizational roots. I, therefore, compare official participation records across these localities after matching them by a series of time-variant covariates within a difference-in-differences approach. The results provide a causal estimate of the effect of new party entry close to a 1% average increase on electoral turnout, robust to different specifications. This finding advises against the use of electoral participation rates as an explanatory variable in studies of new party entry, while it confirms previous findings on new party entry effects. More broadly, it contributes to the growing literature on the effects of supply-side changes on political behaviour. From a normative perspective, it also defies the conventional wisdom that increasing party system fragmentation and volatility have mostly a negative impact on indicators of democratic quality.