The supposedly imperialistic character of government agencies is a vast oversimplification. Autonomy is valued at least as much as resources, because autonomy determines the degree to which it is costly to acquire and use resources. High autonomy means the agency has a supportive constituency base and a coherent set of tasks that can provide the basis for a strong and widely shared sense of mission. If autonomy can reasonably be assured, then the agency of course will seek more resources or an enlarged jurisdiction.
But that is a big "if." As the variety of government activities has increased, the opportunities for any agency to have an uncontested jurisdiction and a wholly supportive constituency have shrunk. Turf problems were not major problems when the only important federal agencies were the Post Office, the Pension Bureau, the army, and the Customs Service. Turf problems are large, and largely insoluble, when the government has within it dozens of agencies that make foreign policy, scores that make or affect economic policy, and countless ones that regulate business activity and enforce criminal laws.
This may explain what John Wanat found in his study of the patterns of growth in various federal agencies between 1952 and 1966. It turns out that after allowing for inflation many agencies did not grow at all.* Those that did grow were disproportionately research agencies.41 Many research agencies (one thinks of the National Institutes of Health) have a very supportive constituency (who would oppose finding a cure for cancer?); an undisputed jurisdiction (there are no other agencies doing this job); and a coherent sense of mission (the research interests of professional scientists). Under these circumstances agency pressures for growth are likely to be strong because they are not costly. Of course, whether the agency's desire for growth actually results in growth will depend on many things beyond its control, including prevailing economic and budgetary conditions.
Very few federal agencies have the advantages of a cancer research institute. Most must struggle for both autonomy and resources. How that struggle is managed by agency executives is the subject of the next chapter.____________________