In recent years, some 40-45% of US merger and acquisitions transactions have involved divestitures of individual product lines, divisions or subsidiaries. The conventional wisdom holds that units with poor earnings records and future prospects are divested more frequently than others, and there is stronger evidence that their purchasers tend to be in the same business sector. In Discussion Paper No. 665, Research Fellow Margaret Meyer, Paul Milgrom and John Roberts develop a model of `influence activities' to account for divestitures. Unit managers' attempts to influence the distributive results of organizational decisions generate real costs: resources devoted to them; value lost when they lead to sub-optimal decisions; and degradation in organizational performance as policies, decision processes or organizational structures are changed to limit influence activities.

Meyer, Milgrom and Roberts note that mergers create new opportunities to transfer rents among units, which may increase influence activities and hence entail costs that can offset the coordination or other gains from the merger. Changes in business conditions can also affect the costs and benefits of a unit's influence activities, the corresponding influence costs and hence its optimal ownership configuration. If a unit's prospects are relatively poor, members undertake influence activities to protect their jobs. This can be limited by isolating the endangered unit so that it can no longer compete for corporate resources. This may be impossible so long as the parent firm remains liable for the unit's debt, so isolation may necessitate divestiture. Influence attempts are also related to the risk of lay-offs and competition for resources within the firm, which may be reduced by selling the divested unit to a firm that can absorb the shrinking unit's employees into its other operations. Declining and unprofitable units should therefore be over- represented in divestitures, and acquiring firms should tend to be those already operating in the same business of the divested unit.

Meyer, Milgrom and Roberts test these ideas formally with a series of models and find that influence activities to protect jobs are uniquely associated with potentially declining businesses; but they do not generally increase with the likely rate of decline. For units facing rapid decline, the rents and quasi-rents associated with continued employment and hence their incentives for influence activities may be reduced, while marginal changes in the parent firm's estimate of the unit's prospects may have less effect on employment decisions, so that the returns to influence activities will be smaller at the margin.

Organizational Prospects, Influence Costs, and Ownership Changes
Margaret Meyer, Paul Milgrom and John Roberts

Discussion Paper No. 665, June 1992 (AM)