Specific Knowledge and Divisional Performance Measurement

This classic by the formulators of agency cost theory discusses five common divisional performance measurement methods—cost centers, revenue centers, profit centers, investment centers, and expense centers—while providing a theory that attempts to explain when each of these methods is likely to be the most efficient. The central insight of the theory is that each method offers a different way of aligning decision-making authority with valuable "specific knowledge" inside the organization.
The theory suggests that cost and revenue centers work best in cases where headquarters has good information about cost and demand functions, product quality, and optimal output mix. Profit centers—defined as business units whose managers have responsibility for overall profits, but not the authority to make major capital spending decisions—tend to supplant revenue and cost centers when line managers have a significant informational advantage over headquarters and when there are few interdependencies (or "synergies") between divisions. Investment centers—profit centers in which unit managers are allowed to make major investment decisions—tend to prevail when the activity is capital-intensive and when it is difficult for headquarters to identify the value-maximizing investment strategy.
In evaluating the performance of profit centers, rate-of-return measures like ROA are likely to be effective when unit managers do not have major influence over the level of new investment. But, in the case of investment centers, Economic Value Added, or EVA, is likely to be the most effective single-period measure because it is designed to encourage only value-increasing investment decisions.