Administration

Administration is often viewed as one of the major overheads of central government. The mean number of administrators per 100 inhabitants for OECD countries (0.30) and developing countries (0.29) is remarkably similar (Table 13). Typically, African countries appear to have the highest burden of administrative costs (0.29 per 100 inhabitants) and Asian countries the lowest (0.14).

Within the OECD, for the countries for which data are available, apparently the countries with the largest administrative sector are Sweden, New Zealand, Iceland, and Ireland. Indeed, 8.75 percent of the total adjusted central government employment in Ireland is represented by administration.

A characteristic of the international comparison of government employment in administration is that those countries that have been more influenced by the British Commonwealth system of government record higher proportions of central government civil servants in administration than do other countries; for instance, Kenya (28.8 percent of central government civil servants in administration), South Africa (19.4 percent), Zimbabwe (14.4 percent), the Bahamas (10.7 percent), and Jamaica (11.6 percent). This might be explained by the fact that the generalist tradition in the British Commonwealth leads many jobs to be classified as administrative that in other countries would be thought of as specialized.

Those employed in the administrative sector of government appear to be paid rather more than the average for the public service as a whole. In OECD countries this ranges from a differential as low as 1 percent in Canada to 41 percent in the United States. What is noticeable is that relative payment to administrative officers in developing countries appears to be higher than in the OECD countries across a wide range of African, Asian, and Latin American countries. It might be questioned whether the elitism suggested by these figures is matched by the (of course, difficult to judge) output.

Education and Health

The proportion of adjusted central government employees involved in education is dramatically higher in the OECD countries than in developing countries; the mean for the OECD is 38.8 percent, compared with 28.5 percent in the developing countries. The country with the highest commitment in terms of the share of employees allocated to education appears to be Belgium (58 percent) and the lowest in OECD European countries, Denmark (21 percent). In developing countries, the figures for Asia and Latin America tend to be higher than those for Africa, and again encompass wide variations, for example, 36 percent of the adjusted central government personnel in Kenya are employed in education but only 12 percent in Zimbabwe. Argentina devotes almost 50 percent of its central government employment to education.

The number of employees per capita in the education sector of the OECD countries is almost three times that in the developing countries; for health, the ratio is four times larger in the OECD region. For both sectors, employment is considerably higher on a per capita basis in the Latin American region than in Africa or Asia.

Where there are extraordinarily low figures for public health employment, they may represent other quasi-official ways of providing health care outside the budget payroll, for example, through insurance (Australia) or lotteries (Ireland). Employment in the health sector in OECD countries averages 19 percent of total adjusted central government employment, compared with 12 percent for the developing countries. The highest public personnel commitment to health is found in Sweden (46 percent) and Iceland (44 percent).

There is considerable variance in the health employment statistics, owing almost wholly to the institutional issue discussed above. In the OECD region, health employment per capita ranges from less than one employee per 1,000 inhabitants in Ireland and the Netherlands, to 53 and 42 employees per 1,000 inhabitants in Denmark and Sweden, respectively. Yet the difference in employment in the total health sector of these countries is far less marked.

Payments to those employed in public sector education appear near the average for the OECD countries, but it is striking how payments to educators in the public service in Africa and Asia are markedly above the average, for example, Zambia (146), Korea (222), and Sri Lanka (148). In Latin American countries, the range is closer to that in Europe. These figures conceal the actual responsibility for paying the salaries of teachers or health workers. In some countries, especially those associated with the French system of government, teachers are hired, fired, and paid by the central government. In other countries (e.g., the United States, the United Kingdom, and Kenya), the teachers are employed by local authorities.

On the whole, public administration theory would support the idea of local influence over the provision of locally concerned public services; in practice, as local authorities usually rely on the central government for substantial transfers of revenue, their ability (even where they have the authority) to determine their own salary scales and hiring practices is limited. The outcome is often the worst of both worlds with localities telling teachers how they would like to pay more but how they are frustrated by their central government’s constraints. while the central government talks about the irresponsibility of local authorities and their poor appreciation of the requirements of broader national macroeconomic policy. No matter how the supposed responsibility for education and health wage decisions is allocated, the public sector commitment to education and health can be crucial for setting wage patterns, particularly differentials, in the urban labor market. Indeed, in 1982 the British Government fought the longest strike in the history of U.K. public sector labor relations over the pay of workers in the state health service precisely because it considered it had to hold down wage awards to contain inflation.

Defense and Police

Defense and police forces should probably be taken together, as the distinction between the two in a crosscountry comparison is likely to be of questionable validity.³² The share of government employment in both defense and police is higher for the developing countries than for OECD countries. In the former, 21 percent of the central government labor force is likely to be committed to defense and 12 percent to police, whereas the similar figures for OECD countries are 15 percent and 7 percent, respectively. The country with the highest percentage of its population committed to defense in the OECD countries is the United States (1.4 defense personnel per 100 inhabitants and 0.4 police). The country with the highest defense and police commitment is Cyprus (2.8 defense personnel per 100 population and 0.6 police).

Initially, it appears that there is little difference between the number of defense employees per capita in the OECD and developing countries. However, if one adjusts for the effects of Singapore, Korea, and Cyprus, the contrast between the developed and the developing countries in the number of defense employees per capita becomes clearer, with the developing countries employing twice as many defense personnel per capita. The variance in the number of defense personnel in those countries is quite stark. One wonders why Swaziland requires 6.9 military per 1,000 inhabitants and Kenya only 0.8 (when their police forces are comparable). In contrast, the number of police per capita does not appear comparable in the developing countries and in the OECD region. Latin American countries appear to employ more police than do countries in Asia and Africa. The variance among OECD countries in the number of police is quite small.

For the countries for which defense data are available, there is no systematic pattern that defense employees are paid markedly lower salaries than those in the other functional sectors. This finding may simply reflect the weight of civilian employees in the defense sector, but only in part. Examining specific countries, defense forces in some OECD countries appear to be paid substantially below the norm (e.g., Japan, 81 percent), while some developing countries pay their defense forces substantially higher amounts than the average for the public service as a whole; for example, Kenya, 130 percent, Zambia, 134 percent, and Argentina, 140 percent.

If defense forces should be considered in conjunction with the police force, then, on average, the police and defense forces receive comparable pay in the sample of developing countries; in the OECD, police are better paid, although this fact probably reflects the effect of lower salaries for draftees in the defense forces of OECD countries. In many developing countries, the police forces appear to have an average wage that is much lower than the average wage of the military, for instance, in Sri Lanka, Argentina, the Bahamas, Ecuador, Zambia, and Jamaica. However, there are examples of the opposite relationship, notably in Swaziland and Korea. It is difficult to appreciate why the pay of the police force should be markedly different from that of the military; many of the duties appear comparable and while some of the skills needed by the military may justify higher pay (for example, pilots), the day-to-day duties and responsibilities of police work might seem to justify a somewhat higher rate of pay in general.

Other Sectors

In terms of the number of employees per capita in finance and planning, there is almost no difference between developed and developing countries. There does not appear to be much difference in the mean number of employees in agriculture per capita between developed and developing countries, but again there is a wide variation between regions, with the mean for Africa being approximately 0.19 per 100 inhabitants, the mean for Asia 0.05, and that for Latin America 0.17. The ratios of central government employees in mining, manufacturing, and commerce or in transportation and communication do not reveal many significant differences between developed and developing countries. There is, however, as might be expected, a significantly higher number of central government employees in the area of labor and social security in the OECD region.

Finally, a much higher fraction of central government employees in the developing countries are engaged in economic services, finance and planning, agriculture, mining and manufacturing, and transport and communications. Approximately 19 percent of employees in the central government are in these sectors in developing countries, as opposed to approximately 9 percent in the OECD countries.