Scholars have long been concerned about the effect that digital inequalities might have on marginalized populations. Concern for the "digital divide" extends to social movement scholars, who worry that the digital divide will lead to social movements privileging the concerns of the middle class over those of disadvantaged groups. We argue for a novel way of testing for such effects—the use of a Heckman regression model to model participation in online activism. The Heckman model separately models selection effects (i.e. first-level digital divides that affect Internet access) and main effects (i.e. second-level digital divides and classic predictors of micro-mobilization). We find that the digital divide in access does not exert a selection effect and that the digital divide in usage exerts minimal effects in models predicting online petition-signing.