Click here for search results

Resources

Subnational Government

Subtopics on this page:

Subnational civil services are a substantial portion of government workers around the world. Excluding health and education workers – which can easily distort the picture if one is interested in administration – subnational  employment is often over 50 percent of the total in federal counties; and with few exceptions, subnational employment is larger than the subnational share of public expenditure (see Table 1). Many of the government services that are key for poverty alleviation are delivered and managed by subnational administrations; and the worldwide trend is toward greater decentralization. For all these reasons, the quality, management, financing, and performance of the subnational civil service matters greatly.

Table 1: Subnational Share of Overall Public Spending and Public Administration Employees (excl. health and education workers)

-

subnational share in public expenditure

subnational share in government administration

-

subnational share in public expenditure

subnational share in government administration

Angola

24.9

14.6

India

53.3

49.4

Argentina

43.9

76.3

Indonesia

14.8

29.4

Armenia

5.1

82.2

Ireland

30.7

39.7

Australia

47.9

63.3

Italy

25.4

48.9

Belarus

32.5

22.1

Kenya

3.5

18.2

Belgium

11.8

57.0

Lithuania

22.6

37.3

Bolivia

36.3

11.3

Malaysia

19.1

31.8

Botswana

3.8

23.8

Netherlands

26.1

24.7

Bulgaria

15.7

0.3

New Zealand

10.8

48.8

Canada

49.4

63.7

Norway

37.4

38.0

Chile

8.5

34.3

Poland

22

60.3

China

55.6

93.3

Portugal

11.6

31.9

Croatia

12.1

28.9

Russia

37.6

33.6

Denmark

54.5

65.2

South Africa

49.8

44.4

Estonia

22.4

34.1

Sweden

36.2

56.5

Finland

41.2

78.0

Switzerland

49.3

54.3

France

18.6

44.2

Thailand

9.6

41.2

Germany

37.8

87.0

UK

27

63.5

Hungary

23.7

47.5

USA

46.4

72.5

Source: Sciavo Campo et al (1997), WDR 1999/2000;

Recent thinking on decentralization emphasizes the role of incentives in the performance of decentralized government. These incentives are shaped by a host of factors, including the electoral and political system at national and subnational level, the intergovernmental fiscal system, the subnational fiscal system, and the (lack of) an active civil society.  The emphasis here is on the civil service management system, though this cannot be regarded as completely separate from the other systems.

Why have a separate subnational civil service?              

There are several good reasons for a separate subnational civil service. Generally, these are an extension of the potential gains from decentralization:

  1. better knowledge of local demands;
  2. ability to respond to local cost variations;
  3. increased scope for community participation; and
  4. closer correspondence between costs and benefits.

However, these improvements are likely to occur only with an executive accountable to the local electorate. Full accountability can only be achieved with a subnational civil service separate  from the central civil service.

Still, there are reasons to maintain a degree of central control over subnational civil services. First, in the management of the civil service there are clear economies of scale in tasks such as training, personnel management, wage negotiations, and the like. Second, in many developing countries, there are scarcities of skills. Some remote areas may not be able to attract the right quality of civil servants without offering prohibitively expensive incentives. They would therefore benefit from some degree of centralization, which would at the same time improve the chances of a more equitable delivery of services across the country. Third, the center often takes an interest in the size of the civil service for budgetary reasons since it may end up paying the bill for a bloated subnational civil service. Finally, the civil service can be a powerful tool for nation building and unity, and some central influence on the decentralized civil service could therefore be desirable, especially in countries where national unity is a concern.

Along with these economic reasons, historic and cultural factors play a large role in determining the degree of independence of subnational civil services. In many developing countries with a colonial past, some key features of the system are still similar to those of the former colonizer, despite the fact that the colonizing country’s system may have evolved considerably since that time.

What are the institutional arrangements around the world?  

Many OECD and decentralized developing countries have a separate civil service for each level of government. However, in unitary countries such as China and Korea, all civil servants are—at least nominally—part of one civil service (see Burns 1997; World Bank 1997; Young-Pyoung 1997).

The Web page on The Scope of the Civil Service in OECD and Select CEE Countries presents data for OECD countries and select Central and Eastern European countries. In this sample of 34 countries, subnational government employees (excluding education, health, and police) are defined as national civil servants in 18 cases. Meanwhile, in 11 cases there is a separate civil service for subnational government. (Click here to view the data table.)

China regularly rotates senior civil servants from one province to another, and from the central ministries to the provinces and back. In federal countries like India and Malaysia, each state has its own civil service, but federal civil servants are sent to work for regional governments on a rotation basis, often in the most senior posts. In countries such as Germany and Indonesia, the subnational civil services are obliged to perform central tasks as an agent of the central government (co-administration), thereby ceding some of their independence from the center. Subnational tasks financed through earmarked grants have similar effects, with the center exercising at least some influence over the subnational civil service. Other unifying elements are common training for national and subnational civil servants, national standards on service delivery, and the education level of particular categories of civil servants (e.g., teachers, doctors).

An analytic framework to determine central authority over subnational civil services                                                             

To illustrate, in many countries the center finances the subnational wage bill directly or indirectly (either through earmarked grants or general transfers). But in developing and emerging economies that are unitary countries, it is usually the center that determines what revenues subnational governments can raise, even if the later may have flexibility in choosing the rates. Thus, the center largely determines subnational budget envelopes. National laws often determine performance standards and (unfunded) mandates for subnational civil servants.

To determine the degree of central authority over subnational civil services, several aspects need to be reviewed, including:

  • Legislative authority. Are all civil servants covered by one overarching legal framework?  Or do national laws only apply to the central civil service, with separate by-laws passed for each of the subnational civil service? Or, has the central government passed a framework law for subnational civil servants within which subnational governments have considerable freedom to determine specific conditions in their region?

  • Structure and Size. Does the center have the power to dictate the structure of subnational government directly? If so, how effective are these controls? Are there significant numbers of government workers outside the civil service, and are their numbers controlled? Does the center control most of the resources to finance the subnational civil service, or do subnational governments have sufficient resources of their own and the ability to decide on the optimal size of their civil service?

  • Career Management. Does the center retain influence over (senior) appointments and career movements? Is there a central civil service secondment system? Are there minimum standards in terms of education, experience, and background for civil servants in general? Is there a central job classification system? Do sectoral laws or regulations define those standards for specific groups of civil servants? Is the center in control of mandatory training for certain positions?

  • Performance management. Does the center determine wage structure and wage increases, or leave it up to the regions within their budget envelopes? Is the civil service solely accountable to the local executive, or does the center supervise? Are there many national service standards that the subnational civil service must meet?

In each of these areas, there is a range of central authority over subnational civil services that range from complete central authority to complete subnational authority. This range is set out in Table 2.

Table 2: Central authority over subnational civil service

Model 1
Total Central Authority

The national and subnational civil service is unified, with no significant subnational distinctions. The consequence is that there are no significant differences between civil servants placed in national or subnational government.

Model 2
Central Dominance

A subnational civil service exists, but the center sets legislation that tightly defines and prescribes its workings. In consequence there are minor differences between central and subnational civil servants, with possibility of some variation prescribed by the center.

Model 3
Central Guidance

The center sets standards and other parameters in legislation covering the subnational civil service, giving considerable discretion to subnational government. In this model significant differences can emerge between the center and subnational government, and between subnational governments themselves.

Model 4
Central Leadership

The center shapes subnational civil service structure and practices through reference power, i.e. customary leadership, professional influence and demonstrated best practices. In this model, significant differences can emerge between center and subnational government, and between subnational governments

Model 5
Autonomous

Subnational governments set their own civil service policies. The consequence is that the potential exists for wide variation between central and subnational government, and between subnational governments themselves.

This analytical framework is tentatively applied for selected countries in
Table 3.

Table 3:  Application of the Analytical Framework
in Selected Countries

  USAUKFranceIndonesia

Policy and Legislation

4

2

4

2

Structure

 -   Establishment control

5

3

4

1

 -   Employment framework

5

3

4

2

Career Management

 -   Appointment

5

2

4

2

 -   Mobility

5

2

4

2

 -   Training

5

3

5

2

Performance Management

 -   Pay and rewards

4

3

n.a.

2

 -   Standards

4

3

n.a.

2

 -   Accountability

5

2

2

2

Source: WB staff assessment (Manning, Hofman, Dice, Campos, Nankivell, Monaghan); Key: Model 1 Total Central Authority; Model 2:Central Dominance; Model 3: Central Guidance; Model 4: Central Leadership; Model 5; Autonomous.

Transition issues in decentralization                                    

Once a country begins to decentralize, several additional issues beyond the structure of the civil service must be dealt with. With the transfer of tasks and resources that decentralization entails, the civil service will have to be transferred as well. If that fails, the central government is likely to suffer higher deficits, or government services may be seriously disrupted. One example of this is Colombia. In the course of decentralization a substantial amount of tax revenues was transferred, but the central government was forced to continue to pay the wage bill for staff in key sectors such as health and education, and suffered a higher deficit as a result. The transfer of civil servants can become more difficult if ethnic tensions exist. For instance, the central civil service in Indonesia is dominated by Javanese, which, apart from their ethnic background, bear the additional stigma of having been associated with the previous Suharto regime. Now that the country has started to decentralize, these tensions are coming to the surface, and placement of central civil servants in the regions has become a sensitive issue, especially for the managerial ranks. Resistance to the central civil service may also arise because of a loss of patronage opportunities at the local level.

The sheer logistics of decentralizing the civil service should not be underestimated. Reassignment of staff may require a number of administrative steps, including the transfer of personnel files, a rearrangement of the payroll, and, sometimes, individual transfer decrees. Civil servants may also have to move from the national capital to provincial capitals or municipalities newly responsible for the tasks they perform.

Aside from these issues, the performance of the civil service in decentralization may suffer because of an initial lack of accountability. Decentralization entails new lines of accountability and reporting—often to an executive that is not used to monitoring the services provided by them. Without diluting subnational authority, the central government can assist by continuing to provide performance information across subnational governments. Before decentralizing the civil service and its  tasks, the central government can also build up independent monitoring mechanisms such as parent-teachers associations, health boards, and various civil society watchdogs.

Solving these issues, and setting up monitoring mechanisms, requires a significant amount of planning and time. Therefore, decentralization is ideally a phased process, starting with political decentralization, followed by administrative decentralization sector by sector, followed by fiscal decentralization. In the process, mechanisms can be worked out to resolve conflicts among levels of government, and between central government and the civil service.

Recommended readings                                                

  • Burns, John P. 1997. "The Civil Service System of the People’s Republic of China." Paper presented at the conference Civil Service Systems in Comparative Perspective, Bloomington, April 5-8.
  • OECD (Organization for Economic Cooperation and Development)/PUMA. 1999. "Managing Accountability in Intergovernmental Partnerships." Public Management Service, OECD, Paris.
  • OECD. 1997. Managing Across Levels of Government. Paris: OECD.
  • World Bank. 2000. "China: Managing Public Expenditures for Better Results. Public Expenditure Review." Washington, D.C., May.
  • Young-Pyoung, Kim. 1997. "The Korean Civil Service System." Paper presented at the conference Civil Service Systems in Comparative Perspective, Bloomington, April 5-8.

This page was authorized by Bert Hofman of the World Bank.  It was submitted on 14 August 2000.

 




Permanent URL for this page: http://go.worldbank.org/1BK5XL6JV0