Capital intensity and the federal sector: Some further evidence

Concluding remarks

The above considerations provide certainly no more than a small contribution to the puzzles of the economics of public bureaucracy. Nevertheless, the empirical evidence presented may reflect some general features of the provision of public services; still, it would be desirable to get additional information from other countries in order to carry positive theory further.

The evidence of a high level of real property intensity certainly asks for an explanation. Most helpful in this respect would probably be a thorough analysis comparing particular service domains of the public and of the private sector in order to find out whether comparable services are provided at a similar degree of real property intensity. Since such comparisons are not to my knowledge available (as yet) it may be appropriate to list some speculative arguments that may serve to explain real property intensity. Several typical features of the provision of public services are presumably of particular importance.

Public administration mainly consists of providing services; in most cases such services are produced — spontaneously or upon request — by public bodies themselves. It is noteworthy and of relevance for real property intensity that the public authorities (can) only carry out relatively few duties ‘off the premises’ (e.g. police inspections, control of national frontiers), whereas the bulk of public services are produced ‘on the premises’. This requires, apart from other provisions, adequate buildings with office space, court halls, waiting rooms etc., in some countries also schools, hospitals a.o. The essential difference between the public sector and many private service companies is probably, that (under the prevailing standards) the public administration can hardly reduce its requirements of real property by going directly to the customer/client in providing its services or even by contracting external professional (who do not require an office) to any large extent.

Another characteristic of the public administration may also contribute to the high level of real property intensity. The input of real property must be seen in the light of well-known claims for public services to be generally and easily accessible to the public and for official authorities to work in close contact with their clients. For many public responsibilities, therefore, there are local and district offices and subsidiaries (e.g. legal courts, local authorities) spread out over the whole administration area, apart from the economically useful concentration — regional as well as functional — in ministries and central offices usually located in the capital. This obligation, in reality probably quite strong, for the public administration to be present at the local level (also for higher-ranking authorities) and to de-centralise personnel and capital may yield positive external effects for clients of public authorities (in the form of lower private costs), but at the same time overincreases factor inputs, particularly also the input of real property. To some, albeit smaller, extent the high real property endowment of the public sector may also be determined by the well-known practice to use for administrative purpose, preferably and independent of cost considerations, buildings which are public property rather than rented office space.

The empirical evidence available on capital intensity is yet too scarce for a test of the ‘factor-supplier pressure group hypothesis’ (Tullock, 1977) for the domain of public administration.¹ Tullock assumes that factor input is determined in different ways on the federal and the local level by civil servants and factor suppliers. The specific electoral interests and bargaining power of these groups suggest that the central administration will be more capital intensive in its provision of services than the local authorities which will rather give preference to the factor ‘labor’. It is not possible to settle this issue with evidence being available only for the federal administration level; yet it can be argued that, abstracting from institutional conditions, the diagnosis of a high capital intensity on the federal level is indeed consistent with Tullock's hypothesis.

To the extent that no other, hitherto not identified factors further narrow the scope for discretionary action (e.g. higher quality standards of services provided at the federal level) the above figures reflect the room for manoeuvre — in the sense of the Tullock hypothesis — within which the bureaucrats successfully imposed their factor-related preferences in the respective domains.

The test by Dilorenzo (1981) using data on local government spending is itself not sufficient since the evidence only refers to the local level.

Further evidence on factor intensity in other domains of the public sector is given in Borcherding et al. (1982).