Stylized facts indicate that small firms are responsible for a disproportionate share of innovative research. There are many possible explanations for this facto The paper seeks to understand this phenomena as the outcome of an optimal assignment of tasks across individuals and organizations. It is shown that incentive costs associated with a given task depend on the total portfolio of tasks that an individual or an organization undertakes. Mixing, hard to measure activities (innovation) with easy to measure activities (routine) is particularly costly, since it will either lead to misallocation of attention across tasks or to misallocation of risk. Larger firms are at a comparative disadvantage in conducting highly innovative research, because of the costs associated with managing a heterogeneous set of tasks. It is further argued that optimal organizational responses to coordination and control of routine tasks will lead to bureaucratization within the firm and to financial constraints imposed by capital markets, both of which are hostile to innovation.