VI. Conclusions This study documents and quantifies a mechanism that has the potential to bias the outcome of shareholder votes by generating extra votes for each share on loan over a vote-record date. 18 The study is particularly timely as Over-voting is currently under regulatory review and has thus far been the subject of scant empirical work. I explain how the over-voting could create bias and then provide empirical evidence that the mechanism is working as described. In particular, I show that 10 shares on loan generate 8 extra sets of voting instructions. This magnitude follows intuitively from an average vote turnout of 88%. I then identify particular events which were mostly likely impacted by over-voting 44 of which remain suspect as inclusion requirements become more restrictive. Finally, I provide evidence that over-voting produces less efficient vote outcomes, which disappoint shareholders and result in firm value destruction. Within the group of 44 vote cases, a one percentage increase in a firm's outstanding loan position is associated with a 2.5% decrease in the firms meeting date return.