The labor market and economic stabilization in Zambia

Policy, Planning, and Research 1 WORKING PAPERS Macroeconomic Adjustment and Growth Country Economics Department The World Bank November 1989 WPS 222 The Labor Market and Economic Stabilization in Zambia Christopher Colciough Labor markets are assumed, in orthodox stabilization programs, to be flexible. The experience in Zambia proved this assumption unfounded. T'he Policy, Planning. and Research Complex distributes PPR Working Papers to disseminate the findings of work in progress and to enoourage the exchange of ideas among Bank staff and al others interesjsj in development issues. These papers carry the names of the authors. reflect only their views, and should be used and cited accordingly. The findings, interpretations, and conclusions arc the authors' own. hey shouid not beattnbuted to the World Bank, its Board of Directors, its management, or any of its membercountries.  Policy, Planning, and Researchi| Macroeoonomic Adjustment and Growvlh This paper examines trends in the Zambian labor Real wages did fall somewhat over the following market over the period since independence. It decade, but insufficiently so to promote eco- focuses particularly on two phenomena - skill nomic diversification and recovery. The rigidi- shortages and wage rigidities -- which have ties also help to explain the failure of more made it more difficult for the economy to recent stabilization efforts, including L IMF recover from the fali in the price of its main program of 1985-87. export commodity, copper, in the mid-1970s. TIhis paper is a product of the Macroeconomic Adjustment and Growth Division, Country Economics Departmcnt. Copies are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Raquel Luz, room NI l- 057, extension 61588 (40 pages with tables). The PPR Working Paper Series disseminates the findings of work under way in the Bank's Policy, Planning, and Research g Complex. An objective of the series is to get these findings out quickly, even if presentations are less than fully polished. !The findings, interpretations, and conclusions in these papers do not necessarily represent official policy of the Bank. Produced at the PPR Dissemination Center  TABL! OF CONTENTS Page Introduction 1 Economic Change in Zambia 1970-85 1 Aspects of Labour Supply 5 Aspects of Labour Demand 9 Labour Market Rigidities 11 Impact of the 1985 IMP Stabilization Package on the Labor Market 30 Conclusion 32 FOL 3tes 36 Appendix 37 References 39 Table 1 3 Table 2 6 Table 3 7 Table 4 10 Table 5 12 Table 6 13 Table 7 15 Table 8 17 Table 9 19 Table 10 20 Table 11 22 Table 12 24 Table 13 27 Table 14 28 Table 15 29 Appendix Table 1 38  Introduction Conventional stabilisation and adjustment policies for developing countries typically presuppose the existence of labour market flexibility. Variation in the nominal exchange rate is usually given a central place in such reforms because it can change the structure of relative prices in virtuous ways. In the presence of labour market rigidity, however, such price changes will not be achieved. If, for example, real wages were inflexible downwards, reduction of the nominal exchange rate would have scarcely any real effects: except during a short transitional period, relative prices would remain unchanged and resource reallocation would not occur. Inflexibility of the labour market can also explain a lack of output response even where relative prices have been successfully altered. Low supply elasticities often arise from a lack of available physical inputs - raw materials, equipment, intermediate goods - but they can also be caused by shortages of or qualitative deficiencies in human resources, by ignorance of technological possibilities or of market opportunities. These facts are demonstrated in this paper, which examines the relationship between successive attempts to stabilise the economy of Zambia and trends in its domestic labour market. It is shown that both skill shortages and real wage rigidity exacerbated Zambia's economic decline over more than a decade, notwithstanding successive nominal devaluations of the Kwacha. Furthermore it is shown that the early demise, in 1987, of the adjustment package imposed by the IMP resulted from an unrealistic (or unconsidered) assumption that the majority of middle and lower income urban Zambians would tolerate pauperisation at the hands of a rapidly depreciating nominal exchange rate, notwithstanding the lessons and history of the previous decade. It follows that serious analysis of the labour market is a necessary input to (and the major past omission from) the design of stabilisation and adjustment programmes if they are to stand much chance of either economic or political success. The analysis begins with a short account of Zambia's macroeconomic record since 1970. This is followed by an assessment of some of the main influences on and characteristics of labour supply and labour demand over the period. The central section of the paper identifies the major labour market rigidities in Zambia and their contribution to the country's economic decline. The final section examines the impact of the 1985 IMF stabilisation package on the labour market, and the reasons for its ultimate rejection. Conclusions are drawn for future policy in Zambia, and for other countries contemplating - or being urged to introduce - economic reforms. Economic Change in Zambia 1970-85 In recent times there have been few countries in the world which have had a more catastrophic economic record than Zambia. Over the two decades ei,ding in 1985, only six countries - from amongst the 120 or so for which data are available (World Bank 1987: Table 1) - had a worse record of income loss. Furthermore in four of these six cases (Niger, C-had, Uganda and F,icaragua) the occurrence of natural or political disasters in the form of lrought and/or war largely explain their demise. 'Easy' explanations are  -2- absent in Zambia's case. Its recent economic history therefore has some important, and unusual phenomena to explain. As argued elsewhere, the major causes of Zambia's economic decline lie in the interaction between the prolonged fall in export prices, and the particular ways in which the Zambian Government (GRZ) reacted to the payments deficits which this price trend caused (Colclough 1988). There were, nevertheless, aspects of the evolution of the labour-market in Zambia which prompted (or prevented the alleviation of) Zambia's difficulties. Table 1 summarises important aspects of recent economic change. During the early 19709 output growth was steady. But, at about 4 per cent per year, it was more modest than during the heady expansion of the immediate post-independence years. In 1975, howpver, growth was arrested. The first part of Table 1 shows the performance of each of the main branches of the economy over the period, expressed in 1970 constant prices. It can be seen that agriculture did continue to grow after 1975, but, at around 2 per cent per year, this was not as fast as population growth (see below). Thus, per capita agricultural production declined over the period. The picture in other sectors appears to have been even worse, with the output of mining, industry and services each remaining roughly unchanged, or falling slightly over the following decade. The immediate reasons for this lie in the performance of the mining sector. These reasons are somewqhat hidden by the constant price production series for mining shown in the table. Since that is basically a volume index - converting quantities to values on the basis of prices obtaining in 1970 - it hides any value losses, or gains, accraing to Zambia over the period as a result of changes in the real international price of its main export commodity, copper. Such price changes have, in fact, been substantial. Most notably this was so in 1975, when the price of copper on the international market fell by almost 50 per cent. At that time the value of copper exports represented about half of Zambia's GDP, and over 90 per cent of exports in value terms. Tl is, the price fall had an immediate, and massive, effect upon Zambia's current income. Table 1 shows that the terms-of-trade loss in 1975 amounted to more than one quarter of GDP in comparison with prices obtaining in 1970. The whole of this loss arose from the collapse of the copper price. The most immediate impact of these disastrous events was felt on the balance of payments. From a fairly comfortable position in 1974, both current and capital accounts swung sharply into deficit the following year. That on current account amounted to approximately 30 per cent of GDP. There were few alternatives to financing the deficit by Zambia increasing its iiaternational borrowing and by running down its reserves of foreign exchange. The latter were sufficient to cover only about 10 per cent of the deficit on current account. Accordingly there was heavy recourse to borrowing as a short-term response. This, of course, carried the implication of higher debt-service commitments over the longer term, which could be met only if the external earnings/expenditure imbalance were to be quickly corrected.  Table 1: Gross Doestic Product and Income (total and per capita), by Type of Acanomic ActivitY (millions of constant 1970 Kwacha) 1970 1974 1975 1980 1983 1984 1985 Agriculture 136.1 150.5 157.0 166.4 172.3 181.9 198.7 Mining 460.3 474.3 427.9 398.7 446.5 402.8 381.4 Industry 235.0 339.4 34r.0 326.8 346.7 349.7 366.4 Other 437.1 509.7 508.2 525.9 506.9 500.8 513.0 GDP (market prices) 1268.5 1473.9 1438.1 1417.8 1472.4 1435.2 1459.5 Plus T/T adjustment(a) - -104.7 -387.2 -367.3 -427.1 -350.0 -360.0 GDY(b) 1268.5 1369.2 1049.9 1050.5 1045.3 1085.2 1099.5 Population (000's) 4159 4700 4846 5647 6285 6500 6725 GDP per capita (K) 305 314 297 251 200 191 190 GDY per capita (K) 305 291 217 186 166 167 163 Notes: aRough estimates for 1984 and 1985 bGross domestic income: this is equal to GDP plus adjustment for variations in the terms of trade. Sources: Calculated on the basis of data in World Bank 1984, Tables 2.02 and 2.04 and World Bank, 1986, Tables 2.02 and 2.04.  a 4 - A range of policies were utilised by the Zambian Government in its atte=pts to achieve such a correction. Gross domestic expenditure fell by 18 per cent in 1976. Th's was necessary if more domestically produced goods were to be released for export, or, alternatively if imports were to fall. However, owing to the structure of the economy, export growth required substantial product diversification, rather than mere contraction of domestic dem'rd. Thus the main impact of the latter was felt on the side of imports which fell by one quarter, in real terms, in 1976. The composition of the expenditure reduction, however, gave additional cause for concern. Almost the entire brunt of the fall in spending -das concentrated on investment expenditure, whilst consumption continued to rise in absolute (though not in per capita) terms. This pattern was sustained during the following several years, which drastically reduced the possibility of a return to economic growth, even fromi the depressed levels of the mid 1970's. Spending by the Government was also cut back sharply. Here again, however, the burden of the cuts was focussed upon the capital, rather than the recurrent budget. Capital spending fell by 70 per cent over the eight years ending in 1983. In spite of this the public sector deficit continued to amount to approximately one-third of total spending. This was partly financed by borrowing from the central bank, and partly by foreign borrowing, both of which tended to aggravate the size of the external payments deficit. As regards policies focused directly upon the external sector, protection was sharpened by the introduction of an elaborate system of import licensing and of foreign exchange allocation. As has been found in many other country contexts, however, this system of controls was administered often inefficiently, and sometimes corruptly.It certainly served to reduce the level of expenditure on imports, but it did so in ways that further undermined the productivity of the domestic economy,owing to the shortages and delays which it helped to generate. The exchange rate was gradually depreciated after 1975. By 1982 it had fallen to approximately two-thirds of its level in the mid-1970's. The pace of domestic inflation, however, was faster than in Zambia's main trading partners and, in spite of nominal adjustments, the currency in fact became increasingly overvalued. The prevailing price incentives, therefore, tended to decrease the profitability of export or import-substitute production over the period. Iu response to the high, and rising inflation rate, the Government increased the administrative regulation of consumer prices, and extended the range and extent of subsidies on basic commodities. Although this protected consumers, it further reduced the profitability of domestic industry. The consequence of the above range of domestic policies wan that the economy continued to stagnate over the period 1976-1983. By the early 1980's, capacity utilisation in manufacturing had dropped to as little as 30 per cent, whilst transport and marketing bottlenecks prevented the effective distribution of those commodities that were available. Continued over-valuation of the exchange rate provided few incentives for diversification. Although imports were held back, the sustained decline of the copper price implied a continued reduction in real export revenues, and, therefore, in the capacity to import. The rising price of oil exacerbated the  difficulties. The real value of exports (adjusted for changep in import prices) was halved between 1974 and 1975 - but it was hdlved again by 1983. The impact of these trends is shown ir. Table 1: domestic income remained stagnant between 1975 and 1985, and by the latter year, income per head had fallen to a level which was scarcely mr.a than half that of 1970. Meanwhile, current account deficits continued, vh!ch were financed by borrowing. Both loans and attendant debt service grew sharply, and by 1983 the latter was equivalent to more than half of export earnings, exclusive of arrears equal to about two years of current debt service payments. An already disastrous situation thus became untenable. A major injection of external resources was needed, which would be forthcoming from the country's main donors only in return for a sharp change in the direction of domestic policy. Such changes were, in fact, introduced, following extensive negotiations with the IMF. The most fundamental of these came in late 1985. Their centrepiece was the introduction of a weekly auction of foreign exchange, which engendered a massive fall in the value of the Kwacha and a sharp readjustment to the whole structure of relative prices. Nevertheless, the reforms proved to be short-lived. A later section of this paper examines the reasons for this, together with the impact of these recent events upon the labour market. Aspects of Labour Supply Population Growth and Age-Sex Composition: Against this gloomy background of economic decline there have been important changes in the characteristics of the labour force over the past two decades. The population had, by 1987, more than doubled since 1963. Table 2 provides estimates to 1980. It shows that its rate of growth has itself increased sharply over the period - from 2.5 per cent per year in the early years of independence, to 3.3 per cent over the period 1974-80. One of the major consequences of the acceleration in the rate of population growth is a change in its age structure. On the one hand it is becoming more youthful: the proportion of the population under 15 years increased from 46 to almost 50 per cent between 1969 and 1980 (Table 3), whilst the proportion in the working age group decreased. At the same time, because of declining mortality rates the proportion of elderly people increased over the period. Table 3 shows that similar trends in the age structure have occurred for both sexes - although if anything the decline in the proportion of males in the working age-group has been sharper than that for females. These trends have had a number of important consequences. First, even if individual fertility declines, birth rates will remain high for a considerable number of years by consequence of the age structure. Second, the number of new entrants to the labour force will continue to increase sharply well into the next century. Third, the decline in the proportion of people ir. the working age-group implies an increase in the dependency ratio, with consequent needs for proportionately higher health and education expenditures, as well as heavier demands on the household budget, at a time when the economy has been in sharp decline. The interaction of these trends has undoubtedly resulted in extreme hardship being experienced by a large number of individuals and families.  Table 2: Pulation Growth and its Distribution, 1963-1980 Average Annual Growth (%) 1963 1969 1974 1980 63-69 69-74 69-80 Total Rural 2774484 2864879 3014000 3239389 0.5 1.0 1.1 Total Urban 715020 1192116 1663000 2440419 8.9 6.9 6.7 of which - Lusaka 123146 262425 401000 538469 13.4 8.9 6.8 0 Kitwe 123027 199798 251000 314794 8.4 4.6 4.2 Ndola 92691 159786 229000 282439 9.5 7.4 5.3 Total Zambia 3490170 4056995 4677000 5679808 2.5 2.9 3.1 Percentage Urban (%) 20.5 29.4 35.6 43.0 Sources: Population Census Results from each of the years shown  7 Table 3: Poulation bY Sex and Ace Grou. 1969 and 1980 (tercentaaes) Age Group Male Female Total 1969 1980 1969 1980 1969 1980 Under 15 years 47.5 50.7 45.6 48.7 46.4 49.8 15-44 years 37.9 36.3 42.9 40.0 40.5 38.3 45-64 years 12.1 9.6 9.5 8.7 10.8 9.1 65+ years 2.5 3.3 2.0 2.6 2.3 2.8 Source: GRZ 1980a, Table 2.3  - a - Internal Migrationt The physical location of the population has also changed markedly. Analysis of the two most recent census results shows that rural-rural migration, primarily within provinces outside the line of rail, has been numerically the most important type of population movement. Nevertheless, rural-urban migration has remained very important - albeit with important shifts in volume over the years since independence. Urbaniqstion has been extremely rapid, such that by the mid 19808 Zambia was the most urbanised country in Sub-Saharan Africa (exluding South Africa), with almost half of the population living in urban areas by that time (World Bank 1987, Table 33). The movement to the towns was greatest in the early post-independence years. Table 2 shows that urban growth of 9 per cent per year in the 19608, slowed to about 7 per cent by the early 1970s. Further reductions in the annual rate of growth to around five per cent, appear to have occurred. Although the most recent disaggregated data are for 1980, ^ is probable that urban growth slowed sharply after that date as a result of the increased economic hardships associated with urban living. Data for the three largest towns are shown in Table 2. Their growth was particularly marked until 1974. Subsequently, the growth of the two copper-belt towns of Kitwe and Ndola fell beb id that of other urban centres. The capital city, Lusaka, however, continu*i to Srow rapidly, at least until 1980, - albeit at a somewhat slower pace than during t..e first independence decade. There has, therefore, been a redirection of migration away from the Copperbelt - which was initially worst hit by the recession of the late 19709 - towards the capital. Net-migration to the Copperbelt fell in relative terms by almost 50 per cent between 1969 and 1980, whilst that to Lusaka changed little over the period. Already by 1980 there wes evidence of increased return flows of migrants from the Copperbelt to Northern and Luapula provinces. Such out-migration from the urban areas most affected by economic decline will almost certainly have further increased over the past few years. Schooling: The provision of schooling for Zambians has increased enormously since independence. Access at primary level has quadrupled, such that by 1984 more than 1.2 million children were in primary school. Secondary school places had increased ten-fold to 125000 by the same year, and university and technical college enrolments had increased likewise to approximately 10000 students. Even such rapid expansion of schooling as this, however, takes many years to change the educational profile of the population significantly - particularly in the context of Zambia's high rate of population growth. Thus, the population census of 1980 revealed that almost half of the population aged 15 years and above had not completed lower primary schooling, and were thus probably still illiterate. Moveover, only 10 per cent had had some post primary schooling, wi-ih tiny proportions having proceded to the tertiary level. These fact ,elp to explain the continued skill shortages which, as will be shown below, represent an important constraint upon Zambia's social and economic development. Participation in the Labour Force: Somewhat unusually, the 1980 census defined the labour force to include all persons aged 12 years or more. They comprised 3.3 million persons. Of those, just over half (1.7 millions) were regarded as being economically active, with the remainder being students, housewives, retired or disabled persons, etc. Using these conventions, 71 per cent of the male labour force was recorded as being economically active, with  -9- only 37 per cent of the females being so classified. For the usual reasons these estimates result in a significant understatement of the amount of productive work undertak % by women. Half of the working population were engaged in agricultural work (two-thirds of the females and about 44 per cent of the males). One third were self-employed and a further 5 per cent (but 12 per cent of the females) were unpaid family workers - again, preponderantly in the agricultural sector. Aspects of Labour Demand Data on the number of employees in the formal sector are shown in Table 4. As is to be expected from what has been said earlier, experiences in the labour market bafore and after 1975 were very different. The first decade of independence was characterised by steady growth of formal employment at about 4 per cent per year, such that, by 1975, approximately 50 per cent more workers were employed than had been the case at independence. Subsequently, however, employment fell back, declining, on average, by about one per cent per year to 1984. This circumstance of absolute decline in formal employment over the decade beginning in 1975 is the most fundamental feature of ZaMbia's recent labour market history. Its effects have been momentous. The proportion of the labour force employed in the formal sector has fallen by one third since 1975. An even more startling implication is that whereas one in thirtieen Zambians had a job in the formal sector in 1975, ten years later only approximately one in 18 persons had such a job. Thus, the number of persons without access to any (direct or indirect) support from wage or salary incomes is now vastly greater than it was in the early 1970s. The structure of formal employment by industry is indicated in Table 5. A number of important features emerge from this table. First, job losses after 1975 were heavily concentrated in the construction industry, which io3t more than half of its jobs by 1984. The primary cause of this concentration was the massive fall in investment expenditures, which resulted from the curtailment of domestic expenditures in response to the country's growing payments deficit. Second, employment also fell in mining and quarrying, particularly after 1980, in response to the sharply curtailed profitability of the copper industry. Thirdly, apart from agriculture, which stagnated over the period, employment elsewhere actually rose: in manufacturing and transport by about 10 per cent, and in services, by about 7 per cent to 1984 (but by somewhat more to 1980). More detailed analysis of the data for services reveals that the rise to 1980 was almost entirely due to expansion o. employment in the public sector, which could not be sustained, for fiziancial reasons, beyond that date. Nevertheless, it remains true that the most severe employment problems were contained in a rather narrow industrial range: outside mining and construction job losses were not so severe. This Lonclusion is strengthened if one considers only Zambian employees. Rapid progress with localisation was achieved after 1975 - that is, an increasing number of Zambians took over jobs previously held by expatriates. Thus, whilst the total job losses between 1975 and 1984 amounted to about 30,000, almost 20,000 expatriate workers appeared to have been  10 Table 4: Emlovees in the Formal Sector by Citizenship, June 1964-1984 (a) Total Employees Zambians Non Zambians Non Zambians as % of Total 1964 264,100 232,000 32,100 12.2 1965 301,600 267,760 33,840 11.2 1966 313,360 284,480 28,880 9.2 1967 303,730 274,650 29,080 9.6 1968 319,730 290,8'J0 28,930 9.0 1969 321,460 291,640 29,820 9.3 1970 337,750 310,320 27,430 8.1 1971 358,350 331,670 26,680 7.4 1972 364,740 338,750 25,990 7.1 1973 377,640 342,990 34,650 9.2 1974 386,270 351,190 35,080 9.1 1975 398,840 365,330 33,510 8.4 1976 379,400 346,310 33,090 8.7 1977 372,630 345,320 27,310 7.3 1978 369,310 344,480 24,830 6.7 1979 371,670 348,290 23,380 6.3 1980 381,490 360,980 20,510 5.4 1981 373,720 355,220 18,500 5.0 1982 367,510 350,050 17,460 4.8 1983 363,800 348,270 25,530 4.3 1984 365,190 351,190 14,020 3.8 Notes: Prior to 1972, data are available for Africans and non-Africans on:y. These are shown in the relevant citizenship columns. Data for 198;-84 are preliminary estimates. Source: GRZ 1985, Table 6.  - 11 - localised over that period. Thus, Zambian employment fell by the somewhat lower figure of 10,000 workers over those nine years (Table 5). The relationships between output and employment growth are investigated in more detail in Table 6. Employment/output elasticities are shown for the 1970-75 and 1970-80 periods. These estimates are rough, in the sense that they take no ar.count of changes in real and relative factor prices over the period. In addition, in a number of sectors the elasticities are not shown, owing to the occurrence of a negative value for one of the relevant variables. Nevertheless, for the economy as a whole the data suggest that the aggregate relationship between employment and output growth has been close to unit;. This implies that in the aggregate, no labour productivity growth occurred over the 19708. This unusual result can be explained in a number of ways. First the elasticities shown are heavil'y influenced by movements in the services sector, which itself is dominated by the public service. Since, here, output increases are basically determined by increases in the real wages and salaries paid, output and employment can usually be expected to move upwards or downwards at similar rates, in the absence of significant changes in the real value of average earnings. Second, circumstances of negative growth of output and employment are not conducive to the measurement of meaningful elasticities. For example, there are likely to be strong political and institutional reasons for laying off workers less quickly than may be thought justifiable on profitability grounds. Thus, the ensuing fall in average labour productivity will tend to raise the size of measured employment/output elasticities during times of recession. At the least, some adjustment lag is to be expected in the labour market, which partly explains the relationships between output and employment indicated by Table 6. Labour Market Rigidities An important question which arises is whether or not conditions in the labour market exacerbated the economic decline. Although cause and effect are not easy to disentangle, it does seem that there were some structural characteristics prominent in Zambia which made adjustment to the new economic circumstances more, rather than less difficult to achieve. Skill Shortages. The first of these concerns the skill-structure of the labour force. Shortages of skilled and educated workers have always comprised a major problem for the country's development. During the colonial period African education had not been given high priority. In spite of the comparative wealth of Northern Rhodesia, during most years far more money was spent upon the education of European children than upon all forms of education for the remaining 97 per cent of the population. In addition, racial discrimination in both the public service and the private sector was often explicit (for example, until 1959, legislation prevented Africans from being taken on as apprentices). Thus, the training and experience necessary for advancement was often not available to the African population. By consequence Zambia started out with a major deficit of professional and industrial skills, which was to prove very difficult to remove. Table 7 indicates the changes in the educational structure of the formal sector Zambian labour force that have subsequently oc.urred. The  12 Table 5: Employees in the Formal Sector. 1965-84, by Citizenship and Economic Sectork81 1965 1970 1975 1980 1984 Agriculture,Forestry,Fishing 32700 34610 36100 32630 35400 Zambian 32070 34090 3479n 31580 34660 Non-Zambian 630 520 1310 1050 740 Mining and Quarrying 52360 57640 64750 63070 58470 Zambian 44820 52130 54440 57110 54390 Non-Zambian 7540 5510 10310 5960 4080 Manufacturing 27770 38160 44330 47760 48200 Zambian 24030 35230 41230 45860 46720 Non-Zambian 3740 2930 3100 1900 1480 Construction 45320 68740 71750 43750 33610 Zambian 43100 66220 67790 42070 32570 Non-Zambian 2220 2520 3960 1680 1040 Transport & Communications 14560 22340 22050 23940 24000 Zambian 11870 19720 20490 23040 23340 Non-Zambian 2690 2620 1560 900 660 Services(b) 88350 121480 154510 168150 165510 Zambian 72290 108200 142430 159780 159490 Non-Zambian 16060 13280 12080 8370 6020 TOTAL 261060 342970 393490 379300 365190 Zambian 228180 315580 361170 359440 351170 Non-Zambian 32880 27390 32320 19860 14020 Notes; a - Data are for fourth quarter of each year shown, except those for 1965, which are averages of all quarters, and for 1984, which are for June. Prior to 1975, the citizenship breakdown refers to Africans and non-Africans, rather than Zambian/non- Zambian, respectively. b - Includes utilities, financial and business services, restaurants and hotels and other services. Sources: GRZ 1980b, and Meesook et al 1986: Appendix Table 1.  Table 6: Observed relationships between Outvut and Eloyment, by Sector. 1970-80 GDP(") Average Annual Average Annual Growth Employment/Output Growth of GDP (%) of Employment (%) Elasticities la) 1970 1975 1980 1970-75 1970-80 1970-75 1970-80 1970-75 1970-80 Agriculture(h) 41.5 56.3 58.7 6.3 3.5 0.8 -0.5 0.12 - Mining 460.3 427.9 398.7 -1.4 -1.4 2.4 0.9 - - Manufacturing 129.2 157.6 162.9 4.1 2.3 3.0 2.3 0.73 1.0 Construction 90.3 138.5 79.1 8.9 -1.3 0.9 -4.4 0.10 - Transport 48.8 57.6 56.5 3.4 1.5 ' -0.3 0.7 - 0.47 Services ' 403.8 499.5 554.2 4.3 3.2 4.9 3.3 1.14 1.03 TOTAL 1173.9 1337.4 1310.1 2.6 1.1 2.8 1.0 1.08 d.91 Notes: a) Millions of Constant 1970 Kwacha, market prices b) Excluding subsistence output and employment c) Includes import duties less imputed bank service charges d) Calculated by dividing average annual growth of employment by that of GDP, for each period and sector Sources: GDP data: CSO, Konthly Digest of Statistics, various issues Employment data: Text, Table 5  - 14 - Hanpower Survey which was held shortly after independence (GRZ 1966), documented in some detail the extent of the manpower constraint faced by Zambia at that time. It can be seen from the Table that the vast majority of the 236,000 Africans then employed had had little or no formal schooling, and that less than one per cent of them had had five years of secondary schooling or more. Over the following two decades, however, the educational expansion initiated soon after taking office by Kaunda's United National Independence Party was to alter this educational profile sharply. Table 7 shows that by 1983 over 20 per cent of Zambian employees had at least 'O' level (secondary. Form 5) qualifications, and that about half had had some secondary schooling. These were major achievements for a country which, twenty years earlier, had inherited a school system which was so thoroughly inadequate in both quantitative and qualitative terms. Nevertheless, this progress fell far short of meeting the needs of the formal sector for the full range of technical and professional skills, and a substantial number of non-citizen skilled workers had to be employed throughout the period. This was in spite of the fact that rapid localisation of the labour force was assigned high political priority soon after independence. A fairly complex set of regulations was established, requiring the satisfaction of strict criteria before work permits for expatriate personnel could be issued by the Ministry of Labour. Foremost amongst these was the production of satisfactory evidence by an employer that no suitably qualified or experienced Zambian was available for the job in question. Under these circumstances, a good indicator of the extent to which skill shortages amongst the Zambian labour force have remained important is provided by trends in the employment of non-citizens. Table 4 shows that the total number of non-Zambian employees remained approximately constant from 1964-76, although their proportional importance declined from 12 per cent to 9 per cent of formal employment over those years. Thereafter, however, the number of expatriate workers dro,pped sharply, such that by 1984 only 14,000 (less than half the earlie) levels) remained, accounting for less than 4 per cent of formal employment. This aggregate picture of good progress with localisation is, however, misleading in a number of ways. First, in one important sense, the 'gap' is even smaller than it appears in Table 4, in that approximately half of the non-citizens currently employed are Zambian residents, mainly originating from other African countries, who, from a labour market viewpoint, can be treated as being equivalent to citizens since they are unlikely to face official pressures to leave.3 On the other hand, in a different sense, the progress with localisation has been less satisfactory than the table appears to indicate: although large numbers of non-citizens left Zambia after 1975, not all of those workers were replaced by adequately qualified and experienced Zambians. Many of their jobs remained vacant. This was particularly so for those expatriates who vacated professional and technical jobs after 1985 and who left the country primarily because they faced declining real incomes in the face of the Kwacha devaluations which occurred. By 1987, the situation in the health sector was particularly serious, with large numbers of foreign doctors having left Zambia, without replacement by either Zambian or expatriate personnel (see below). In addition, the aggregate quantitative picture hid qualitative problems in the labour market which affected most  1, Table 7: Educational Qualifications held by Zambian Employees, 1965 and 1983 a) 1965 1983 Number Number % Less than Form II 226681 96.0 189514 54.4 Forms II/III 7282 3.1 85123 24.4 ;0' Level(b) 1516 0.6 45308 13.0 A' Level(b)/Diploma 517 0.2 17566 5.0 Degree 150 0.1 10759 3.1 TOTAL 236146 100.0 348270 100.0 Note: a - Data for 1965 refer to Africans, rather than Zambians. Those for 1983 are based upon a sample survey of employees along the line of rail, covering approximately 70 per cent of total formal employment. The percentage distribution revealed by that survey has been applied to separate estimates for total Zambian employment, published by the CSO. b - '0' level and 'A' level are examinations typically taken after 5 years and 7 years of secondary schooling, respectively. Sources: GRZ 1966, Table II.3, GRZ 1983, Table 8.0, and GRZ 1985, Table 6.  - 16 sectors of the economys localisation had not always and everywhere been efficiently handled. Thus it is evident - particularly in the public and parastatal sectors - that key posts were sometimes filled by Zambians who had had insufficient training or experience. This problem became particularly acute as regards deficiencies in management skills - often at quite senior levels of the public service. Finally, those non-citizens who remained in employment were often concentrated in particular specialised or technical areas, requiring long periods of high-level training. Thus the process of localising the remaining jobs held by expatriates appears more problematic than has typically been the case in the past. Equally, the premature departure of such workers would generate problems that would be commensurately more difficult to solve. Data on the characteristics of those in wage or salaried employment in Zambia tend to be unreliable, and inconsistent as between different sources. It is,. therefore, difficult to build up an accurate picture of the educational, sectoral and occupational composition of non-Zambian employment. Howeve- Table 8 shows the occupational breakdown of total employment as revealea by a recent manpower survey, which is of interest given the comparatively high coverage of the formal sector which it achieved.4 The table shows that, although non-citizens were, by 1983, a comparatively small proportion of total employment - less than 5 per cent - they were nevertheless proportionately much more important in professional and technical occupations, where they comprised 12 per cent of employment, and in administrative/managerial jobs where they accounted for more than one-fifth of total reported employment. It can be seen from Table 9, which shows all occupations with more than 130 non-Zambian workers, that some occupations in the professional group were still heavily dependent upon expatriate skills. The case of engineers is particularly startling, where more than 1000 expatriates were employed, comprising almost half of all such workers. Zambia also remained heavily dependent upon expatriate doctors. According to the survey, they comprised more than three-quarters of all doctors in the country (although this proportion is probably correct, the figures are significant under-estimates of the absolute number of doctors who were employed). Other occupations where at least one-quarter of all jobs were occupied by non-citizens include physical scientists, statisticians, accountants, university teachers, and managers. These data suggest, therefore, that there was in 1983 a significant training task ahead in order to localise all of these professional posts. At first sight it may seem surprising that non-citizens still occupied a large number of jobs in the 'production, transport and related' occupational category: they comprised almost 4,200 workers, and accounted for one-third of all non-citizens employed (Table 8). Inspection of the more detailed tabulations from the survey reveals that, numerically, the largest single occupation in this category is miners (736 non-Zambian workers). Nevertheless, it can be seen from Table 9 that there were also substantial numbers of non-citizen production supervisors and foremen (597), machinery fitters and assemblers (435), electrical fitters (177), plumbers and welders (169), bricklayers and carpenters (259), transport equipment operators (425) and other labourers (741). Although the proportional importance of non-citizens in these occupations is modest - usually less than 5 per cent of  17 Table 8 SIile of Emlovees by Maior Occuvational Group and Nationality, 1983 Occ. Group Zambian Non-Zambian Total (2) as % (1) (2) (3) of (3) Professional, technical & related 37,676 5,285 42,961 12.3 Admin. & - managerial 2,563 744 3,307 22.5 Clerical 39,925 816 40,741 2.0 Sales 4,583 321 4,904 6.5 Service 35,580 752 36,332 2.1 Agriculture, Forestry, etc. 16,021 527 16,548 3.2 Production, Transport & related 114,886 4,188 119,074 3.5 Not stated 900 35 935 3.7 Total 252,134 12,668. 264,802 4.8 Estimated % of total Employment 72 82 73 Sources: GRZ 1983; GRZ 1985.  - 18 - total workers - these data do indicate that there remained a significant shortage of Zambian artisans in many of the highly skilled craft occupations, both in the mining sector and in other areas of the economy. It should be recalled, however, that approximately one half of all non-citizen workers have long-term residence status, and are not 'targets' for localisation in the same way as are other non-citizen workers. The manpower survey indicated that ovar 4,000 of the 6,000 or so expatriates with work permits, as opposed to residence permits, were in the professiontl and technical occupational group. This proportion (70 per cent) compares with only 15 per cent of those with residence permits who fell into this occupational group (Table 10). This occupational concentration is also to be expected on the basis of the educational background of non-resident expatriates - 55 per cent of whom were reported to have university degrees, with a further 20 per cent possessing professional qualifications of various kinds. (GRZ 1983: Table 10.7). Thus, most of those non-Zambians shown in professional jobs in Table 9 were expatriates who could be localised as more qualified Zambians become available, whereas those with long-term resident status tended to work in the lower-level craft, primary teaching, clerical, farming and nursing occupations. Thus, the recent sharp declines in the number and proportions of non-citizens employed are not so strongly indicative of an increased availability of Zambian skills as they might first appear. Considerable numbers of expatriates have left the country owing to the declining purchasing power of the Kwacha. Those who have left the public sector have often not been replaced - sometimes because of an inability to attract new staff, and sometimes because of a wish to focus the impact of expenditure cuts on non-citizen, rather than Zambian staff. One result has been a significant increase in the number of vacancies amongst professional and technical grades in all sectors of the economy. Unfortunately, comprehensive data on vacancies in the formal se tor are not available. The 1983 manpower survey did not collect such information, nor do the quarterly surveys of employment and earnings run by the CSO. Nevertheless, interviews with senior managers in all sectors of the economy during 1986 confirmed that an increase in vacancy rates was very widely experienced (UNDP 1986). As for the public sector, more concrete information is available. The civil service grew fairly rapidly by, on average, almost 5 per cent per year over the eight years prior to 1983. Payroll costs accounted, during this period, for a growing proportion of public expenditures. In circumstances of increasing financial stringency, expendl.tures on supporting services, (such as transport and materials) were reduced to levels incompatible with efficiency. In 1984 the government, at the behest of the World Bank and the IMF and as part of the conditions associated with extending new loan facilities to Zambia, agreed to introduce a freeze on recruitment to the administrative cadre of the public service. Thus, a sharp increase in the vacancy rate occurred, and by 1985 it was estimated that some 10,000 of the 77,000 established posts in the civil service were vacant (UNDP 1986:155). In 1986 GRZ went further, in announcing its intention to make no new appointments to the public service without the special authority of the Cabinet Office, and to  19 Table 9: occupations in which Non-citizens are concentrated. 1983 Occupation Total Zambian Non-Zambian (3) as 4I (1) (2) (3) of (1) Physical Scientists 430 292 138 32.1 Architects and engineers 2,476 1,381 1,095 44.2 Engineering technicians 3,222 2,874 348 10.8 Doctors 334 73 261 78.1 Nurses and midwives 4,104 3,798 396 9.6 Statisticians and related technicians 832 625 207 24.9 Accountants 3,531 2,716 815 23.1 University and other Higher Ed. Teachers 1,472 1,081 391 26.6 Secondary Teachers 3,410 2,775 635 18.6 Primary Teachers 14,295 13,809 486 3.4 Managers 2,249 1,573 676 30.1 Stenographers 7,120 6,921 199 2.8 Book-keepers 7,278 7,160 188 2.6 General clerks 15,669 15,430 239 1.5 Cooks 4,602 4,438 164 3.6 Protective service workers 18,362 18,083 279 1.5 Farm managers 1,493 1,345 148 9.9 Agricultural workers 13,348 13,030 318 2.4 Production supervisors/ foremen 8,215 7,618 597 7.3 Miners 15,054 14,318 736 4.9 Machinery fitters 11,345 10,910 435 3.8 Electrical fitters 4,587 4,410 177 3.9 Plumbers, welders 5,841 5,672 169 2.9 Bricklayers, carpenters 9,270 9,011 259 2.8 Material handlers 6,790 6,650 140 2.1 Transport operators 9,846 9,421 425 4.3 Other labourers 29,410 28,669 741 2.5 sub-total 204,585 193,923 10,662 5.2 Other workers included in survey 60,043 58,039 2,004 3.3 Total in Survey 264,628 251,962 12,666 4.8 Source: Calculated from GRZ 1983, Appendix 1.  20 Table 10: Non-Zambian Employees, bY type of Permit and Maior Occupation Group. 1983 Work Residence Entry Not Total Permit Permit Permit Stated Professional, Technical land Related 4,263 944 23 48 5,285 IAdministrative and Managerial 364 362 6 12 744 ,Clerical 127 678 1 9 816 Sales 116 20; 1 - 321 Service 140 580 5 28 752 Agricultural 165 358 2 2 527 Production, Transport & Related 840 3,000 100 249 4,189 Not Stated 11 22 - 1 34 Total in Survey 6,033 6,149 138 349 12,668 Source: GRZ 1983, Table 10.0.  - 21 - cut the daily paid employee work force by 40 per cent (about 20,000 jobs) over a two year period. Data are not yet available with which to review the detailed impact of these measures. Nevertheless, various sources report that they have tended to further undermine the effectiveness of the public service, rather than to improve its efficiency. The problem is that retirement, or contract expiry, does not necessarily occur in a way which reflects manpower priorities. An unstructured approach to reducing public sector staffing via 'natural wastage' may therefore compound the problems of production and service delivery. An example can be given from the Ministry of Health, which, in 1985 undertook a review of medical manpower. High and unevenly distributed vacancy rates for nurses and technicians were causing major difficulties for the provision of health care, particularly away from urban centres. The situation amongst doctors, however, was particularly critical. Table 11 shows how the staffing position developed between 1981 and 1985. It can be seen that although there was a net reduction of 86 expatriate doctors over those years, they were 'replaced' .y an increase of only 41 Zambian doctors. Thus, the vacancy rate rose from 20 to 30 per cent over the period, after allowing for increases in the number of posts in the establishment. Again the shortages varied widely in different parts of the country - with vacancies reaching almost 60 per cent at Ndola hospital on the copperbelt. As it happens, these vacancies were caused more by the deterioration in the value of local salaries paid to doctors than by the freeze. But they are illustrative of the potentially misleading import of aggregate data on localisation 'progress' over the past decade. The shortage of skills remains an important rigidity in the Zambian economy - increasing the costs of economic restructuring, and reducing both its speed and potency. Earnings and Labour Costs. The issue of wages and incomes policy and performance in Zambia has received considerable academic attention in the past, with notable contributions from Knight 1971, Dan.el 1979, Pry 1979 and Meesook et al 1986. These generally showed that the wc-cders in the copper industry had acted as wage leaders for those in other s,^ctors, and that the strength of their organisation, combined with the goverr.mint's willingness to accede to their demands in the early years after indepetdence, had established Zambia as a high wage economy from an early stage. There was lively debate as to the extent to which wage increases presented a 'problem' for the country's development. Jolly and Williams (1971) showed that about half of the increase in the cost of living over the late 19609 could be statistically attributed to increased wage costs. In a period where the economy received increasing windfall gains from the rise in copper prices, such increases in average earnings could be accommodated - even though the competitiveness of industry outside copper was badly affected. There was, therefore, a certain complacency about these trends expressed at the time, and even by more recent commentators (eg Fry 1979a 84-85). Whatever view one took of that early period, by the time of the copper price fall in 1975, the Zambian economy was in no way able to restructure production easily away from copper, into manufacturing or service production. Its wages in the non-agricultural sector had by this time become considerably higher than in most countries of a similar size, structure and income level. This is demonstrated by the  22 Table 11: Staffing Levels of Doctors. 1981-85 Year Establishment Zambian Non-Zambian Vacancies Vacancy Rate % 1981 808 62 577 169 21 1982 808 79 586 143 18 1983 830 92 584 154 19 1984 830 88 531 211 25 1985 850 103 491 256 30 Source: Ministry of Health, Manpower Development Office  - 23 - analysis shown in the Appendix. Data are available for 19 countries which allow a comparison of monthly wage-earnings and per capita incomes. For these purposes the year 1976 has been selected in order to illustrate the relative position of Zambian wage-earners one year after the fall in the copper price. It should be noted that Zambia is not selectively disadvantaged by the particular year chosen for this comparison, since the real average earnings of Zambian employees remained approximately constant between 1975 and 1980 (Table 12). As explained in the Appendix, there are methodological and statistical problems which affect this kind of comparison. Nevertheless, some important contrasts emerge. Zambia is shown to be a high-wage economy. This can be demonstrated in a number of different ways. First, and most obvious, is the question of the absolute value of wages paid in Zambia in comparison with other countries. The Appendix Table shows that average monthly earnings from wages outside agriculture were generally, in 1976, two-to-three times the US dollar value of any of the other six sub-Saharan African countries included in the comparison. Although Zambian wage-earnings were roughly similar to those in Papua New Guinea, Cuba and Costa Rica, the only country where earnings (converted to US dollars) were significantly higher was Venezuela - much richer than Zambia, and with by far the highest per capita income of any country in the sample. The results are similar for agricultural wages. Of the eleven countries for which comparable data are available, only two (Korea and Cuba) appear to have paid higher agricultural wages than Zambia in 1976. The latter were about 20 per cent more than the average for all countries having the relevant data (compared with a 33 per cent differential for non-agricultural wages). As for the ratio between agricultural and non-agricultural wages Zambia appears not to have been unusual. Its wage ratio at 2.62 is fairly close to the mean value (2.49) for all countries shown in column 9 of the Appendix Table. Furthermore, Zambia shares the median ranking with Costa Rica. It is not, therefore, the relative levels of rural and urban wages which have distinguished Zambia from elsewhere, but, rather, the absolute level of wages paid, in all sectors, in comparison with other similar countries. A further way to analyse the earnings data is to compare them with levels of per capita income. This is done in column 8 where the ratios of earnings outside agriculture to per capita incomes are shown. Once again, Zambia emerges well up in the rankings. Only Burundi and Kenya have higher ratios than Zambia. Both of these countries are considerably less well off than Zambia, and, at least in Burundi's case, the outcome is more influenced by the low value for the denominator (per capita income) than by the numerator (earnings) being high. Finally, it is worth noting that all of the countries which were poorer than Zambia in 1976 had a wage structure which was also absolutely lower - mostly by a considerable margin. For the nine countries in this category, the value of non-agricultural earnings in seven of them were half, or less, those in Zambia. Agricultural sector differentials were less dramatic, but remained sharp.  24 Table 12 Average Annual Real Earnings of Employees by Sector 1965-83 (1975 Kwacha)(a 1965 1970 1975 1980 1983 Agriculture, Forestry, Fishing Zambian 327 492 453 518 393 Non-Zambian 4410 6350 2524 2902 1215 Mining and Quarrying Zambian 1570 2179 1478 1668 1186 Non-Zambian 10034 10254 6784 5278 3604 Manufacturing Zambian 924 1133 1179 1063 756 Non-Zambian 5836 7590 5657 5243 2972 Construction Zambian 612 860 764 805 632 Non-Zambian 6549 9694 6868 3847 2850 Transport Zambian 924 1710 1834 1308 1006 Non-Zambian 5743 6651 7377 5182 2662 Services(b) Zambian 869 1119 1246 1146 884 Non-Zambian 5149 5535 4894 4278 2726 Total Zaambian 890 1210 1140 1134 851 Non-Zambian 6476 7207 5572 4601 2897 Notes: a - generally 4th quarter figures. Nominal earnings estimates for Zambians deflated by low-income CPI. Those for non-Zambians are deflated by the high- income CPI. b - includes utilities, trade, hotels, finance and other services. Source: Meesook et al, 1986: Appendix Table 3  - 25 - These results are important from the standpoint of international competitiveness. They do not, of course, decisively indicate the existence of differences in labour costs. Such judgements would reguire additional information upon the productivity of labour amongst the countries in the sample. Nevertheless the differences which have been revealed are sharp enough to require implausibly high productivity lev-els in Zambia relative to other similar countries, if such labour cost differences in the mid 19708 were not to have held. There can be little doubt, then, that the level of wages (and labour costs) were higher than would be easily compatible with Zambia rapidly increasing her export revenues in manufacturing and other 'non-traditional' sectors in order to fill the gap created by lost earnings from copper exports. A reduction in wages would by no means have been sufficient to achieve such diversification, but it is clear from the above comparisons that such a change would have been a necessary component of any serious attempt to respond to the country's changed trading circumstances. The movement of average earnings in real terms for the first two decades of independence is shown, in constant 1975 prices, in Table 12. Average earnings peaked in 1970. But the fall to 1975 was primarily caused by the decline in average earnings.in the copper industry. Wage increases during the 1970s in the mining sector were less than the rate of inflation, and represented a period of consolidation after the very large increases in earnings granted to mineworkers resulting from the Brown Commission in :966. In most other sectors average earnings continued to rise until 1975, particularly in services which - at least as defined in Table 12 - covered a large proportion of total formal employment. By this time, therefore, the mining sector was no longer exerting successful wage leadership over the rest of the economy.5 The picture appeared to change, however, over the following years. By 1983 the average earnings of Zambian employees had fallen to three-quarters of their 1975 level, with sharper declines being experienced by workers in manufacturing, transport and services. These declines in real earnings came from a number of sources, including changes in the structure of the employed labour force, changes in the earnings structure, and movements in nominal earnings which less than fully compensated for inflation. It is not possible, on the basis of available data, to dissentangle the relative importance of each of these variables to the net results summarised in Table 12. It is clear, however, that government policy towards wages and salaries had a more profound effect upon the earnings structure - particularly after 1980 - than had been the case earlier on. This occurred in a number of ways. First, the main instrument which had led to the relative protection of real wages in Zambia had been the exchange rate - which had appreciated slightly over the early 1970s. Since most consumption goods are imported, exchange appreciation reduces, ceteris paribus, the Kwacha cost-of-living. Following the devaluation in 1976, however, the value of the currency continued to move downwards. This gathered pace from 1980 onwards, and by 1983 the Kwacha had been reduced to about half of its value in 1975. It can be seen from Table 13 that the impact upon domestic inflation was substantial, with the rate doubling to 20 per cent per year between 1975 and 1983. These  - 26 - were circumstances in which it was easier to achieve reductions in the real wage and salary bill in all sectors of the economy. Second, although the income decline was suffered by workers throughout the formal sector, the earnings losses were felt particularly strongly by those in the public sector. Between 1975 and 1980, for example, workers in the public sector experienced a 16 per cent decline in real earnings, yet those in the parastatals suffered only a one per cent loss, whilst Zambian employees in the private sector were, on average, 14 per cent better off over the period.6 After 1980, on the other hand, the real earnings of workers in all sectors fell as inflation gathered pace (Tables 12 and 13). Survey data are not yet available to allow discussion of trends in average earnings beyond 1983. Yet it was in later years that the major changes in macroeconomic policies occurred. Although these changes were introduced primarily to deal with the economy's structural disequilibrium on the balance of payments, they had sharp implications for the distribution of incomes in Zambia, and for trends in real earnings from formal sector employment. Again, public sector workers were affected most, although earnings in all sectors were severely reauced. This is illustrated by Table 14 which shows the development of the government's general administrative salary scale (covering all sections of the service) since 1967. The nominal salaries payable have been converted, in Table 15, into real terms, and are expressed as ratios of real salaries in 1975. A number of conclusions can be drawn from this table. First, the government has consistently narrowed the gap between the highest and lowest salaries in the administration over the past twenty years. Top salaries declined sharply in real terms to 1975, whilst the lowest paid employees continued to enjoy real earnings increases until that date. Subsequently, although the real earnings of all government employees declined, the fall was much sharper and faster for the senior officers than for those in junior positions. By 1982 the better paid had lost between 40 and 50 per cent of their 1975 real earnings, whereas earlier losses were made good somewhat for the lowest paid, such that in 1983 they were still only 10 per cent worse off than in the mid-70s. These trends continued between 1983 and 1985, with the real incomes of all except the very lowest paid dropping to well below one-third of their 1975 level. In October 1985 the government implemented a salaries revision. For the administrative grades of the service the increases were in two stages - the first payable in October 1985, and the second instalment payable in June the following year. It can be seen from Table 15 that this review continued the earlier policy of narrowing the income gaps in the salary structure. Nevertheless, even after implementation of the first part of the review, most grades received less than one third the real value of their 1975 earnings. The junior clerical and unskilled groups fared better, but their earnings fell to less than three-quarters of their 1975 level. Moreover, these changes coincided with the introduction of the weekly foreign exchange auction and a resulting 70 per cent devaluation of the Kwacha (see Table 13). During 1986 prices rose by more than 50 per cent above their levels of the preceding year, primarily as a result of the massive increase in import prices engendered by devaluation. By consequence, the additional nominal salary increases that  27 TABLE !3: Average Annual Kwacha/Dpllar Exchanqe Rates and Inflation: Selected Years Urban CPI Annual % change in CPI US $ per Kwacha Low Inc High Inc Low Inc High Inc 1970 1.40 1975 1.55 100.0 100.0 10.1 8.5 1980 1~27 202.9 189.4 11.7 11.9 1981 1.15 231.3 209.1 14.0 10.4 1982 1.08 260.2 236.7 12.5 13.2 1983 0.80 311.2 278.6 19.6 17.7 1984 0.56 373.5 336.8 20.0 20.9 1985 0.37 513.5 446.6 37.5 32.8 1986 July-Sept 0.16(a) 778.4 707.2 51.6 58.4 1987 April(b) 0.05(a) 938.3 921.6 37.4 52.4 May 0.125(a) Notes: (a) From October 1985 onwards the value of the Kwacha was determined at a weekly auction of foreign exchange. Wide fluctuations took place during 1986. The value shown is the average value obtaining in the months July-September. Further deterioration took place. The lowest value reached is that shown for April 1987. As from May 1987 the auction was suspended, and the Kwacha's value was pegged at US$0.125. (b) CPI figures for 1987 are for the month of January. Inflation rate for 1987 calculated by comparing these data with January 1986 CPI, as reported by CSO. Sources: World Bank 1986, Bank of Zambia, and GRZ 1987.  1: GwmiU Aitiattiv g:le: 1967-0614g Scale post am. 67 mne 71 July 75 Jan 79 Jan. 80 Jan. 82 Jan. 83 Nov. 83 Oct. 85 July 86 33 iamr-Sscretsry 6.250 7.600 7,812 7.914 10.176 10,176 10.176 10,176 11,536 12,216 87 Assistant Director 5,120 5,800 6.324 6,468 7.440 7,40 7,440 7,440 8,680 9.300 U12 Oniv. Grad. (gntzfl(b) 2.184 2,340 2,976 3,132 3.852 3,870 3,876 3.992 4.931 5,400 S13 Diploa Intry 1.72S 1,860 2,388 2,544 3.264 3,300 33,336 3,465 4,283 4.692 Sal Lmst FPid Salaried 312 396 540 696 984 1.164 1,324 1,470 2.:46 2,464 bvior Unskilled non-established n.G. 342 480 636 840 1,056 1,236 1.382 2.060 2,400 : h.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. (a) lb dates abn represent dates of lplentation for all salaries coiasboas fra 1967 arwards. (b) k entry polnt mm moved upwardr In 1979. Salaries aomn before that date were not entry points, but wro the equivalent polint co the scale to tUe shwam for later years. ": Ibr scace prior to Octobr 1965, see Neeook at .1, 1986, Table A.l.3: 76. For October 1905 and Jam 1986, see GRZ 1906s.  olaJ 1S: l.ndle cf rdl eMlrl. In the Public Sector: Ganral Mdai.trative Scale 11975 - 100) Scale Poet Jan.67 hNy 71 July 75 Jan 79 Aug. 80 Jan. 82 Jan. 83 'Mv. 83 Oct. 85 July 86 53 Under-Secretar' 135 132 100 60 60 59 5i 45 29 22 S7