A dominant aspect of ownership in the developing world is pyramidal structures, which allow shareholders to control corporations with relatively low investments. The uneasy relationship between these controlling investors and minority shareholders, and the potential impact on the broader macro economy, has been well studied by corporate governance scholars. On the one hand, the mismatch of cash flow and control rights leads to a range of agency problems and resultant resource misallocations, potentially impacting the macro economy (see Randall Morck, Daniel Wolfenzon, and Bernard Yeung 2005 for a comprehensive overview). Yet pyramids are one important mechanism that enables the formation of diversified business groups that, too, are a dominant feature of business organization in much of the world. The economics and management literature has taken a more ambivalent view of business groups, with their agency problems and rent seeking behaviors often counterbalanced by productive efficiencies from correcting market failures in weak institutional environments (see, in particular, a survey of the business groups literature by Tarun Khanna and Yishay Yafeh 2007 which emphasizes this tension). Often, the channels that enable value extraction by controlling interests overlap with those required to overcome market failures. For example, consider a controlling shareholder of a publicly listed firm that has a separate, completely owned, Trade and The InTernal OrganIzaTIOn Of fIrms

We study how sharing a hometown or college connection with an incumbent member of China’s Politburo affects a candidate’s likelihood of selection as a new member. In specifications that include fixed effects to absorb quality differences across cities and colleges, we find that hometown and college connections are each associated with 5–9 percentage point reductions in selection probability. This “connections penalty” is equally strong for retiring Politburo members, arguing against quota-based explanations, and it is much stronger for junior Politburo members, consistent with a role for intra-factional competition. Our findings differ from earlier work because of our emphasis on within-group variation, and our focus on shared hometown and college, rather than shared workplace, connections.