Budgeting as a decision process: The biasing effects of compensation scheme, receiver involvement, and source credibility

One of the critical decision facing firms is how to best allocate limited firm resources across alternative investment choices. The development of an accurate and unbiased operating budget is critical to the attainment of efficient resource allocation and profit maximization. This dissertation models the preparation of the operating budget as a judgment and decision-making (JDM) process made under conditions of environmental uncertainty and investigates three factors that may bias the budgeting process: compensation scheme, receiver involvement, and source credibility. Hypotheses about biases associated with these variables are based upon the predictions of the Elaboration Likelihood Model (Petty and Cacioppo 1986a, 1986b). The decision process is modeled using the Hogarth and Einhorn (HE) belief-adjustment model (Hogarth and Einhorn 1992), which models sequential processing of information as an anchor-and-adjustment process. This study examines the sales forecast as the budget variable of interest and uses the HE model to measure subjects' sensitivities to sales-increasing and sales-decreasing messages within the different manipulated conditions. Research is conducted in an experimental laboratory setting using MBA students as subjects.; As predicted by the ELM, low involvement receivers show evidence of using source credibility as a heuristic, listing more source-related decision arguments than high involvement receivers. In high involvement receivers, the direction of credibility's effect depends on whether the message advocates the receiver's initial attitude, which was manipulated through incentive compensation formulas to produce a bias toward either sales-increasing or sales-decreasing messages. The receiver's compensation scheme, the source's credibility, and the direction of the message jointly influenced the weighting of sales-increasing messages, but had no detectable effect on sales-decreasing messages.; This study also tested the contextual validity of the HE model for budget decisions. The findings validate the contrast effect, but do not support the predicted recency order effect. Recency effects appear to be mitigated or even reversed when subjects enter the decision process with a pre-existing bias due to linkage between the final budgeted variable and compensation. This suggests that the HE model, while descriptive of sequential processing of information under some conditions, may not accurately predict order effects when the decision-maker is biased due to economic incentives.