Numerous attempts to explain the decision making framework of public agencies with multiple goals have been published in recent years. The central problem of the multiple goal decision process of public agencies is the problem of unvalued benefits stemming from projects which the agency could potentially implement. Regardless of the complexity of the problem, the fact remains that choices must be made between projects. As a result, several alternative procedures for analyzing the issue have been suggested. These may be subdivided into three more or less different approaches: (1) estimation of value indexes or weights for the arguments of the objective function, (2) constrained optimization models, (3) political decision procedures. The last alternative provides the setting for the discussion that follows. The object is to suggest procedures for incorporating better information in valued and non-valued benefits into public resource. Agency programs of the United States Forest Service are analyzed in terms of their valued benefits and their impact on regional employment. Given that there are significant dependencies in certain Colorado communities on the timber industry, how important is Forest Service timber-harvesting policy for these industries and communities? This question is the basis for the objectives of this thesis. These objectives may be stated as: 1. to quantify the cost of providing for regional income objectives via timber harvesting and investment, and 2. To determine the distribution of benefits from public timber harvesting-investment programs. Are the benefits accruing to particular local economies, or are they leaking out of the region? A present value model is used to value timber investment alternatives. When present values are negative and it can be argued that the investments were undertaken to provide for timber originating employment objectives. To analyze incidence of benefits, regional ownership of capital resources in lumber manufacturing and harvesting are first identified. The proportion of incremental income generated by timber investments which is allocated to non-resident capital is then identified by analysis of factor shares. Analysis of employment sustained by timber harvesting and investment confirmed the generally accepted conclusion that Forest Service timber policy can be an important factor in regional economies. The present value analysis indicated that timber investment alternatives in Colorado typically yield negative values at reasonable discount rates. The distribution of benefits from Forest Service timber policy is not limited to the local regional economies which are presumably of primary concern. Sixty-one percent of the capital int he lumber manufacturing sector of Colorado was found to be owned by outside interests. It was also found that at least ten percent of the harvesting in the state is carried out with non-local capital. When these results are combined with the finding that approximately twenty-four percent of the benefits or value-added by the lumber and wood products sector are allocated to capital it is seen that fifteen percent of the value-added in lumber manufacture and around two and one half percent of the value-added in harvesting are distributed to non-local individuals whose welfare is probably not a concern of decision makers.