We draw on the positive political theory and campaign finance literatures to examine how interest groups allocate influence activities (e.g., monetary donations, lobbying) across multiple government institutions when seeking more favorable agency policy decisions. By modeling agency behavior in the context of legislative oversight, we derive testable predictions about the political conditions under which an interest group will influence (1) only the agency, (2) the legislature and/or executive instead of the agency, and (3) the legislature or executive in addition to the agency in order to induce a shift in regulatory policy. One implication of our conclusions relating to (2) and (3) is that empirical studies seeking to identify a relationship between electoral campaign contributions and public policy using data on legislative votes are potentially misspecified.