The Ambiguity of Bureaucracy

by John Toye

Is bureaucracy a vital institution that has to be built up by poor countries in pursuit of economic development? Or is it, on the contrary, an institution that persistently threatens, like a fast-growing riverweed, to choke the channels of public administration? To understand the ambiguity of bureaucracy, we need to distinguish between the abstract and the concrete senses of ‘bureaucracy’. It is abstract bureaucracy that bears the negative connotation ,while the concrete noun— bureaucracy, synonymous with an organized civil service — does not necessarily do so. This distinction allows us to imagine that the essential policy problem for poor countries is to design a non-bureaucratic bureaucracy. Before jumping to this conclusion, however, we need to look more closely at what is wrong with bureaucracy in the abstract sense.

Those who use ‘bureaucracy’ as a term of abuse are probably making one or more of five different complaints. The first of these complaints is that officials are accountable only to their superiors in the hierarchy, and not to those whose affairs they administer. Officials are empowered first of all by the prevailing laws, but then, under the law, by their superiors delegating powers and duties to them. This implies no accountability to the governed. 

The second complaint is that a bureaucracy operates without any competition, and in the absence of competition, has no incentive to force down the cost of producing public goods and services.

The third complaint is that to the extent that the bureaucracy is providing regulatory services, it is in danger of being ‘captured’ by the private interests whose activities it is supposed to regulate.

The fourth complaint arises because modern bureaucracies operate by making and enforcing rules that apply to categories of people. The purpose of this practice of making general rules is to eliminate arbitrariness, personal favouritism and objectionable discrimination in administration. However, all such general rules usually have some exceptions, from the point of view of complying with common sense— exceptions that are not foreseen or written into the general rule. Yet officials may apply the written rule literally and exactly, and without the exercise of any judgement and discretion. 

The fifth complaint is the multiplication of offices and departments, which operate without adequate co-ordination. This induces a failure of high-level overall control of the bureaucracy. In these conditions, delegation becomes incoherent, and bureaus operate with overlapping and conflicting functions. As a result, people suffer unnecessary delays while trying to find out which official is responsible for the matter concerning them.

Is it possible then to eliminate these negative features of bureaucracy, and design non-bureaucratic bureaucracies? What are the correctives to these fi ve complaints? Peter Evans (2003) has proposed that ‘the effectiveness of public institutions depends on “hybridity”, an integrated balance among three different (sometimes contradictory) modes of guiding public action’. The three modes are: enhancing bureaucratic capacity, following market signals, and empowering bottom-up democratic participation. Evans’s ‘tripod model’ is depicted in Figure 1.


By the mid-nineteenth century, bureaucracy was attracting criticism because autocratic rulers had successfully subordinated it. Since then, the democratic control leg of the Evans tripod has been strengthened. Yet even elected politicians in long established democracies have to struggle to maintain the upper hand in relation to their bureaucrats. It would be naïve to suppose that with the recent spread of democratic regimes to Latin America, the former Soviet Union, Asia and Africa, that the problem of democratic accountability has disappeared. However, there is a fear in some international organizations that any increase in democratic control could disable a bureaucracy from being effective for development. The World Bank’s East Asian Miracle study thought that East Asian bureaucracies were effective because they were insulated from day-to-day political interference. Yet what constitutes ‘political interference’, and what is the right degree of insulation? When does democratic control stop and political interference begin? These central unresolved issues of modern bureaucracies must continue to be the subject of discussion and the object of political contest.

The market signals leg of the Evans tripod addresses the issue of reducing government inefficiency. That issue is clouded by the difficulty of measuring government output. Nevertheless, some improvement in efficiency can probably be achieved by n-th best measures, such as finding small components of a public service that can be out-sourced, by simulating the conditions of competition where they cannot naturally prevail, or simply by insisting that departments surrender a regular small percentage of their expenditure as ‘efficiency savings’, if only to force them to examine the make up of their current costs and make straightforward economising choices.

Regulatory capture arises because of concentrations of political and economic power that become mutually dependent. In industries where oligopoly prevails, existing firms have an incentive to capture the political power to regulate, as a means of deterring potential new entrants. Political parties have an incentive to promise to provide anticompetitive forms of regulation in return for financial contributions to their operating expenses. The bureaucrats may have an incentive to prefer any type of regulation to a scrupulous insistence on enforcing only regulations that are a genuine public benefit. The pressures for collusion are powerful, and to lessen them once collusion has taken hold would have to involve radical political change initiated from outside the system of collusion.

What of the other two categories of complaint? The implementation of bureaucratic rules will always remain problematic. There is an inherent difficulty in anticipating within the written rule itself all the circumstances under which it might have to be implemented. The attempt to deal with every possible case always increases the complexity of the rule, and this probably reduces people’s ability to understand it. If, on the other hand, the rules are kept simple but officials are granted discretion to interpret them, other problems arise. Some will not use their discretion, while those that do may take different views about what common sense requires in the circumstances. The governed will then be subject to what is sometimes called a ‘post code lottery’, namely that while the rules appear to be the same everywhere, what actually happens in a particular case will depend on the jurisdiction where one lives or where one registers one’s business. The enlargement of official discretion opens the door for the return of personal favouritism in the application of rules. Once permitted discretion is there, the next step is that some officials will start selling their favours to those who pay, fuelling corruption. 

The problem of blurred lines of responsibility is not easy to remedy either. In the short run, one can just demarcate official rights and duties more sharply. In policymaking, campaigns for ‘joined-up government’ can do something to mitigate the follies of excessive departmentalism. In service delivery, there is often scope for organizing a ‘one-stop shop’ at the point of public access. The trouble is that such moves, worthy as they are, can never be once-for-all operations. The management of a civil service must be viewed dynamically

The discussion has shown that the ambiguous evaluation of bureaucracies is fundamental and deeply seated, since measures to address bureaucratic defects are often the source of new problems, and need to be applied on a continuing basis. Thus, the prospects of smart designers producing successful blueprints for a non-bureaucratic bureaucracy are not promising. The task is therefore to maintain eternal vigilance, and to balance continuously the trade-offs between further reforms of each leg of the reform tripod.

See Peter B. Evans (2003) ‘Harnessing the State: Rebalancing Strategies for Monitoring and Evaluation’, mineo.​

angle-2008-1_img11.jpgJohn Toye has been successively a Professor of Development Economics at the Universities of Wales, Sussex, and Oxford. He has also worked for the United Nations, as a Director for the Globalization and Development Strategies Division of UNCTAD Secretariat 1998-2000. He has written seven books, his most recent being The UN and Global Economy. John Toye is the author of Chapter 5, ‘Modern Bureaucracy’ in the WIDER study Institutional Change and Economic Development Edited by Ha-Joon Chang and published by Anthem Press (co-published by UNU Press) November 2007.