Despite the significant positive implications of gender quotas, growing evidence suggests that some women serve as proxies for male family members. What is the magnitude of this phenomenon in patriarchal contexts where men gatekeep women's public presence? We address this question by developing a behavioral measure of "proxiness" that is based on who answers the phone to respond to our survey --- the female politician or their family members. Phone survey data of over 1100 local politicians across two states of India reveal that female politicians are significantly less likely than their male counterparts to answer the phone and participate in our survey themselves. We validate this measure with data from a subset of village councils, where citizens (N=969) were asked to name their elected representative. In the village councils where the elected representatives answered our phone survey themselves, they were also more likely to be identified by citizens as the defacto representative. Additionally, we find that women from families with prior electoral experience were more likely to serve as proxies. This study demonstrates the effectiveness of our proposed measure as a low-cost, scalable tool for identifying proxiness at the local level, while also highlighting the gap between gender quotas and women's substantive representation.