Cost Structure and Sticky Costs

Beginning with Anderson, Banker and Janakiraman (2003), the accounting literature attributes the asymmetric cost response to activity changes or sticky costs as resulting from managerial choices. Using a simulated dataset that excludes any role for managerial actions, we investigate whether cost structure alone could yield regression estimates similar to those the literature interprets as consistent with sticky costs. We show that either of two aspects of cost structure — fixed costs or scale diseconomies in variable costs — results in regression estimates consistent with sticky costs. We employ an algebraic model to offer intuition for these effects and to develop suggestions for how future research might control for the effects of cost structure. For a sample of COMPUSTAT firms, we show that the suggested controls provide an alternate explanation for results usually interpreted as reflecting current period managerial actions.