To control for systemic differences between firms based on observable characteristics, we employed a series of controls. Age of the social enterprise may affect a social enterprise’s ability to develop longer term and more productive relationships and to build resources and capabilities for innovation. We lagged the model by using employment and turnover data from the prior year to capture the effects of size on social innovation activities. Models were run with dummy variables for geographical location (compared to social enterprises operating across multiple regions as the default). We included dummy variables for industry sector (business services and marketing; environment, renewables and energy; education and youth services; health and social care; employment services; retail and leisure; housing; and financial services) but found that industry was nonsignificant so do not include this in the final models. Finally, we include a dummy variable to indicate whether the social enterprise is classed as a ‘‘social firm,’’ a specific category that has a particular remit to employ the disadvantaged and that might affect the results (Ducci et al. 2002).