DRD DISCUSSION PAPER Report No. DRD296 THE LABOR. MARKET IN ZIMBABWE: HISTORICAL TRENDS AND AN EVAWATIO.N OF RECENT POLICY by Peter R. Fallon Hay 1987 Development Research Department Economics and Research Staff World Bank The World Bank does not accept responsibility for the views expressed herein which are those of the author(s) 9\nd should not be attributed to the World Bank. or to its affiliated organizct;t.ions. The findings, interpretations, and conclusions are the results of resc~arch supported by the Bank; they do not necessarily represent official poli~cy of the Bank. The designations employed., the presentation of material, and any maps used in tnis document are solely for the convenience of the reader and do not imply"the expression of any opinion whatsoever on the part of the World .Bank or its affiliates concerning the legal status of any country, territory, city, area, or of its authorities, or concerning the delitnitations of its boundaries, or national affiliation. The Labor Market in Zimbabwe: Historical Trends and an Evaluation of Recent Policy Abstract The paper provides both an examination of recent labor market trends in Zimbabwe and an evaluation of recent policies. The first part supplies a labor market overview, an analysis of employment and wage trends, and a review of the evidence regarding communal agriculture and the urban informal sector. It is argued that the recent repeal of migratory restrictions upon Africans has led to a limited transfer of previously concealed unemployment in communal areas into open unemployment in the towns. The second part estimates the effects of wage policies since independence and concludes that current policy has an overall wage restraining effect although some wage differentials are affected. It is argued that existing policies to expand education and training are a more appropriate method of removing African disadvantage than wage policies or job security regulation. THE LABOR MARKET IN ZIMBABWE: HISTORICAL TRENDS AND AN EVALUATION OF ~ECENrPOLICY Table of Contents Page 1. Overview of the Labor Market 1 2. Formal Sector Employment 5 (i) Sectoral Trends 5 (ii) The Quality of the Formal Sector Labor Force 7 3. Wage Trends and Differentials 9 (i) Institutional Background 9 (ii). Changes in Wages over Time 11 (iii) Wage Differencials by Race 14 4. The Non-Formal Sector 15 (i) Communal Areas 15 (ii) The Urban Informal Sector 19 5. Labor Market Interventions 22 (a) Background 22 (b) Minimum Wages 23 (c) Policy Towards Wage Differentials 25 (d) Wage and Employment Effects of Incomes Policies 27 (e) Job Security Regulation 29 The Labor Market in Zimbabwe - Historical Trends and an Evaluation of Recent Policy 1. Overview of the Labor Market The labor force aged 15 years or more in Zimbabwe was around 2.95 million persons in mid-1987 out of an estimated total population of 8.51 million. Although non-adult labor accounted for more than 3% of the labor force in the 1970~s, workers aged under 15 years have since become much rarer due to subsequent expansion of the educational system. The population is predominantly African (97.5%) with persons of European descent making up about four fifths of the remainder. The European labor force is of disproportionate importance, however, given its much higher level of training and managerial skills. The female labor force participation rate among those fifteen years of age and above is quite high at around 50% although the overall participation rate is unremarkable at 35%, as roughly half of the population are of less than working age. The female participation rate is higher in rural than in urban areas, (52% and 37% respectively). This is the usual pattern in developing countries and relates to the more domestic nature of non-wage rural activities as compared to formal wage employment. Employment is allocated across three broad areas of the economy: formal wage employment - which subdivides further between commercial agriculture and non-agricultural wage activities; employment in subsistence agriculture on the communal lands; and a rather undeveloped informal sector. - 2 - Although Zimbabwe is a predominantly agrarian country with 72% of its labor force in rural areas, this proportion is lower than in most other sub-SaharRn African countries as formal non-agricultural wage activities account for an unusually high proportion of total employment. Sectoral breakdowns of employment in 1969 and 1982 are given in Table 1. The formal wage sector was marginally the largest in 1982, although its increased share is entirely due to rapid growth in the first half of the 1970's. The relative importance of both commercial and communal agriculture has fallen over the period, while the informal sector (both rural and urban) has almost tripled in relative size. Formal sector employment as a whole has virtually stagnated since 1982 and this would indicate that rapid informal sector growth has continued. Historically, the main intersectoral linkage has been via the outmigration of Africans from the communal areas to the non-agricultural formal wage sectors, although a minor proportion of labor employed in commercial farms is supplied by seasonal migrants and residents of neighboring communal lands. The formal wage sector is a relatively well integrated labor market, although, as explained further below, demographic pressures and slow wage employment growth are contributing to a serious and growing unemployment problem. The 1982 Population census estimated an open unemployment rate of 10.8% for the labor force as a whole, and one of 18.5% when communal farmers were excluded. A person was here defined as being unemployed if he or she had done no work but had been looking for work in the previous week. The 1982 estimates represent a huge increase in u~employment when compared with the results of the 1969 census which, under a similar definition of unemployment, found that of the African adult labor force, 4.12% of males and 0.4% of females were unemployed. There is no reason to believe that unemployment has since fallen and government officials informally estimate the current - 3 - unemployment rate at between 15% and 25%. As the census estimates given in Table 2 show, the incidence of unemployment tended to decline with both age and education, but unemployment rates were still absolutely quite high at all educational levels. In most less developed countries, unemployment rates tend to peak at upper primary or lower secondary levels. This was not the case in Zimbabwe, probably because, unlike other African countries, primary enrollment rates fell through the 1970's while secondary rates stagnated. As will be discussed later, minimum wages were at a maximum in real terms during 1982 and this would have placed uneducated jobseekers at a serious disadvantage relative to those with primary schooling~ Male and female unemployment rates were broadly similar and this indicates a lack of labor market discrimination in this respect. The concentration of unemployment among the younger age groups is a common phenomenon in both developed and less developed countries and reflects the initial search period following labor market entry and a greater movement between jobs by younger workers. Concern about the considerable relative educational disadvantage of the African population has led the government since independence to initiate a policy of universal education at both primary and secondary levels. This has been strongly reflected in first year secondary enrollments which rose from 22,201 in 1980 to 82,262 in 1981, and had reached 153,439 by 1985. The average level of schooling of new labor market entrants has thereby increased dramatically very recently. For example, whereas 81% of those in the labor force in 1982 aged 15 to 19 years had only received primary education or less, the majority of new labor market entrants during 1987 will have received between four and six years of secondary education. One anticipates the - 4 - educational expansion to have two effects. The first is that labor market entry that would have been taking place during the early to mid 1980's has been postponed until roughly 1986 onwards, while the small fall in labor force participation that one would expect to follow secondary educational expansion is likely to be offset by higher female participation related to better average qualifications. The strain of increasing labor supply is thus being felt more severely from the present onwards. The second effect that is likely to occur is that there will be a shift in labor supply towards non-manual jobs in the formal wage sector. Here, any slowness in adjustment arising from sticky wage differentials or initially inflated job expectations is likely to exacerbate the unemployment situation for. some years to come. The broad picture that emerges below is that population growth in communal areas and in earlier years, African immigration from other countries, resulted in a relatively elastic supply of unskilled African labor to the formal sector. Legal restrictions on urban residency and a lack of government intervention to support formal sector wages were not conducive to the growth of an urban informal sector or urban unemployment. Any labor surplus would thus have been invisibly concentrated in communal areas. To some limited extent, these conditions have been reversed since independence, which has assisted the growth of visible unemployment and informal activities in the towns. The overall African labor surplus in both town and country has been growing over t~e past decade or so, given stagnation in formal sector employment and rapid labor force growth. Skilled labor is not, however, in excess supply given the net emigration of skilled Europeans during the UDI period and the early years of independence. - 5 - 2. Formal Sector Employment (i) Sectoral Trends There were an estimated 1.06 million formal sector wage employees in June 1985. The coverage of the quarterly wage survey carried out by the C.S.O. in Zimbabwe i's very wide and excludes only wage employment in small agricultural establishments and casual work in the urban informal sector. Temporary and contract workers are included. The distribution by sector for 1985 as given in Table 3 shows clearly that commercial agriculture is still by far the largest employer, followed by manufacturing, domestic servants and public administration. Primary production accounts for 30% of employment, industry for 21% and the tertiary sectors for the remaining 49%. The public sector, i.e. public administration, education and health, now employs 19% of all wage employees, while somewhere around a further 5% are employed by parastatals. Over three quar.ters of all wage employees are therefore employed in privately owned establishments. The overall long-term pattern that one observes between 1955 and 1985 is that agriculture and domestic servants have consistently grown more slowly than total employment; the public services, financial services and other services have consistently grown more quickly; while the growth performance of other sectors has been rather uneven over different time periods. Essentially the long-term employment performance of any given sector reflects a mixture of different effects including: (i) th~ natural tendency for that sector to rise or fall as a share in GOP as the economy develops; (ii) the rate of labor productivity increase due to technical progress; and (iii) the ease with which non-labor inputs can be substituted for labor as real product wages - 6 - rise. Shorter term changes in employment growth over different periods can be further related to the sensitivity of the sector to general domestic cyclical fluctuations and to the timing of internal and external disturbances. In addition to all of these, policies that affect product and factor prices will further influence the sectoral composition of output and the choice of technique within each given sector, while other government policies will directly det~rmine the behavior of employment in the public services. The observed sectoral pattern of employment growth in Zimbabwe arises from a complicated mixture of all of these effects. Taking overall trends first, the long term decline in agriculture is partly due to a natural tendency for the share of that sector's output in G.D.P. to fall as development proceeds and to the introduction of new technique. Agriculture is also a sector in which factor substitution is relatively easy so that farmers have adopted less labor intensive techniques given higher wages. Similarly, the growth of financial services, other services, wholesale and retail trade, and to some extent transport and communications, reflects the high income ·~lasticity of demand that is nor~ally associated with these sectors. The expansion of public services reflects both growing demand for education and health services and deliberate government policy. The unevenness of employment growth in mining, reflects changing market conditions while that of construction and manufacturing reflects in part the sensitivity of these sectors to domestic cyclical fluctuations. The decade following UDI in 1965 was essentially one of unusually high employment growth, generated initially through import substitution encouraged by sanctions. This led in turn to a decade of ~apid output and expansion, particularly in manufacturing and in construction. The - 7 - intensification of the war of liberation between 1975 and 1979 gave rise, however, to a period of overall employment decline, partly compensated in employment terms by a rapid expansinn in public administration due to army recruitment, with direct disruptive effects in the agricultural sector. More disturbingly, however, employment growth has continued to be rather slow since independence, in spite of rapid growth in the number of school t.eachers accompanying educational expansion. If public sector services are excluded, then total employment actually fell at an average of 0.5% per year between 1980 and 1985. The basic pattern has been that employment growth in the two largest sectors - commercial agriculture and manufacturing - has been disappointing when compared with the 1965 to 1975 period, public sector employment has grown very rapidly, while the private service employment sectors - wholesale and retail trade, financial services, and transport and communication - have expanded at their earlier rates of growth. Together, however, these latter sectors only account for 13% of total employment as opposed to the 42% of agriculture and manufacturing. In wage employment terms therefore, Zimbabwe is becoming an increasingly service oriented country with a corresponding decline in the relative importance of the tradable goods sectors. (ii) The Quality of the Formal Sector Labor Force. The majority of the labor force are unskilled, although as Table 4 shows, the skill distribution varies considerably across sectors. The public services and to a lesser extent, financial services, are the sectors with the highest educational content; the private service sectors have proportionately the most skilled and semi-skilled white collar staff; mining, manufacturing, and electricity and gas are the most intensive in sktlled manual workers, - 8 - while agriculture and construction are the most unskilled activities. There are extreme differences in the skill distributions between African and non- African wage employees. For example, using data on skill distributions from the 1981 Manpower Survey and projecting the 1978 CSO estimate forward to 1981 for total numbers of African and non-African workers, one finds that, including domestic servants, 1.7% of African employees were professionals, 6.2% were skilled, 17.8% were semi-skilled and 74.4% were unskilled; while among non-Africans, the corresponding percentages were 21.6%, 34.8%, 19.5% and 24% respectively. This tremendous difference in skill distribution reflects the disadvantaged position of Africans in terms of access to education and training prior to independence. This very striking difference in skill levels runs across all the production sectors with non-Africans occupying a disproportionate number of posts in those sectors with a high skill content. For example, in 1981, there were more non-Africans than Africans employed in the financial services, public administration and education sectors. Non- Africans also accounted for a much larger share of formal sector employment (10.1%) than the labor force as a whole. The emigration of skilled whites has almost certainly led to a decline in the overall quality of the formal sector labor force. Zimbabwe first became a net emigrant nation in 1976, shortly after the intensification of the liberation war. From the beginning of 1980 to the end of 1985 almost 99,000 emigrants were officially recorded as opposed to 40,000 immigrants. These figures al~ost certainly·understate the true size of net emigration as some residents who stayed away permanently or for a period exceeding one year {the qualifying period of absence for an emigrant) had declared themselves to be leaving for a shorter period. The volume of net migration declined sharply - 9 - in 1985 and Zimbabwe became a net immigrant country again in .1986. It is noticeable that the number of residents returning from visits to other countries also increased by 21% in 1985. The skill loss associated with emigration was proportionately most severe with res:p.~ct to the white collar occupations. For example in 1981, 29.3% of economically active emigrants had held professional, technical or administrative posts as compared with 4.4% in the formal labor force as a whole. Emigration in 1981 alone created a loss in these categories equivalent to as much as 5% of higher level formal sector manpower. In skilled blue collar occupations the effects were less acute: partly because emigrants were less relatively concentrated in this group relative to the resident labor force; and partly because discriminatory attitudes prior to independence had tended to classify some African workers as semi-skilled who have since proved themselves to be worthy of a higher classification. The Government has since encouraged this process, both by allowing semi-skilled workers to graduate upwards by passing trade tests and by improving skill levels in the Vocational Training Centers. The expansion of the skill base has, however, been impeded by a falling number of indentured apprenticeships offered by employers. 3. Wage Trenos'and Differentials (i) Institutional Background Wages in Zimbabwe have been subject to a varying range of institutional influences over time. These include (i) government appointed wage setting bodies, (ii) minimum wage laws and, (iii) government prescribed limits to wage increases. In the private sector, the wages of a growing . proportion of workers have been set either by Boards - government appointed - 10 - bodies made up of employers and appointed worker representatives, or by Councils - government appointed bodies made up of employers and trade union representatives. Boards are bodies created by governme.nt within which trade unions are either absent on insufficiently organized, while Councils are the forum for the determination of collective bargaining agreements. The system of Boards and Councils was established by the Industrial Conciliation Act (1959) although some Labour Boards had existed before that time. The system has continued since Independence, although their titles were changed from Industrial Boards and Councils to Employment Boards and Councils under the Labour Relations Act (1985). Both sets of bodies are covered on an industry as opposed to a craft basis, although coverage has not always been complete. In some industries there is duplication with both Board regulations and Council agreements existing in different undertakings. Boards are not a forum for negotiation ..and simply recommend wage increases to government. Collective bargaining agreements are also currently subject to government approval. In 1979, 14.7% of formal sector workers were covered by Boards and a further 15.7% received wages set by Council agreements. At this time, agricultural workers, domestic servants and other private sector employees accounting for a further 12% of formal employment remained uncovered. The Industrial Board for the Agricultural Industry has recommended regulations for agricultural workers since the beginning of 1980, while coverage has since increased throughout the rest of the private sector. At present there are 70 Employment Boards and 20 Employment Councils in operation as opposed to 67 and .21 respectively in 1979. Wages in public administration, education and health are determined by the Public Services Commission. Although European workers in some skills and industries may have - 11 - benefited in the past from a reasonably well developed trade union organization, the state of African unionization was extremely weak at independence. This reflected in part the deliberate use of the Industrial Conciliation Act as a means of controlling the registration and activities of African unions. Some sectors such as agriculture and domestic servants were entirely non-unionized. It was felt therefore that collective bargaining could not be relied upon to improve African wages and to reduce the huge gap between African and European workers. The Government therefore introduced two further policy measures. The first was the setting of a number of national minimum wages in 1980 which covered the entire formal sector and which have since b~en periodically revised, while the second was to directly constrain wage increases as recommended by Boards or awarded by Councils. All workers, including domestic servants are therefore now covered by a minimum wage regulation~ and virtually all workers are subject to a maximum wage increase. In 1986/7 the annual Review of Wages and Salaries took the degree of control one step further by imposing stipulated increases according to income range. Policy since 1980 has thus been consistently one of compressing salary differentials.. Although the minimum wa~re was used during the early 1980s in an attempt to raise unskilled African wages, this role seems to have been abandoned in later years. (ii) Changes in Wages over Time The overall trends in real consumption wages are summarized in Table 5 for agriculture and three other selected sectors. There are three striking features: (i) African real wages grew by much less and at a steadier rate in both commercial agriculture and domestic service than in other sectors; (ii) African real wages grew more slowly during the boom decade 1965/75 than - 12 - in the previous decade, while non-African real wage growth reveals an exactly opposite pattern; (iii) African real wage growth did fall further during the slump years of 1975/8, but a much more pronounced drop in the growth rate took place among non-African workers. The broad conclusion that emerges is that African real consumption wages are much less cyclical than those of non-Africans.!/ This would suggest that not only has African labor supply been growing at a faster rate, but that it is also more wage elastic. The very low growth rates observed in both agriculture and domestic service are consistent with the proposition that the supply price of unskilled African labor has remained virtually constant over time. This fits in with the presence of a 'labor reserve' in the communal areas and the availability of unskilled labor from Malawi, Mozambique and elsewhere in the earlier periods. The rather high observed growth rates in African wages during the 1955-65 period in manufacturing and construction may indicate an improvement in the general skills of the African work force in these sectors, and a growing need in the manufacturing sector for employers to offset labor turnover and its associated replacement training costs with wage premia. There is also likely to have been scme segmentation in the labor market between agriculture and mining on the one hand, and the industrial sectors on the other; given the much heavier use of foreign labor in agriculture and mining, and the seasonality of agricultural employment. The l/ This conclusion is further supported by the following statistical test. Fitting an autocorrelation corrected regression of the logarithm of the real consumption wage on a quadratic time trend and a cyclical variable - here measu1.ad as the deviations of real GOP about its own quadratic trend, one finds that the coefficient on the cyclical variable is positive and statistically significant in almost all sectors for Europeans. For Africans, however, the cyclical effect is never significant. - 13 - generally slower growth in African wages during the decade of 1965-75 may be partly due to a slightly higher rate of growth in labor supply than previously. However, the much less cyclical pattern in African wages observed over the period 1965 to 1978 as a whole, would also indicate that the African labor supply was more elastic than that of non-Africans. This is further supported by a higher observed degree of cyclicality in African employment before 1978. · The ro~e of pre-independe~ce institutional wage determination is less clear. Excluding agriculture and domestic servants, a significant and rising proportion of formal sector workers would have been subject to wage regulation from 1959 onwards. African wages would be subjec~ to more control than those of Europeans who would have d{sproportionately enjoyed any benefits from collective bargaining. The institutional explanation for differing cyclical wage behavior would then be that African wages were fixed according to a simple inflation formula, while those of non-Africans were less regulated and fluctuated accordingly. If this were true, then one would have expected shortages of African labor to appear during the growth period of 1965 to 1975, when wage growth was much less than during the previous decade if institutional intervention were holding real African wages below market clearing levels. There is, however, no evidence of African labor shortage during this period. Similarly, one would have expected institutional intervention to have held African wage growth down during the war and the consequent rise in inflation between 1975 and 1980. Yet although African real consumption wages did fall slightly during this period, Europeans suffered even greater decreases in their real wages, thus generating the rather unlikely conclusion that policy at this time was to help African wage earners - 14 - .relative to Europeans. There seems to be no reason to suppose that the Industrial Board system did much more than to simply recommend wage rates than ' . would have been paid anyway, although the system may have played a role in equalizing African wages both within and across different sectors. (iii) Wage Differentials by Race The average non-African/African wage differential was very large at Independence. In commercial agriculture, for example, non-Africans earned 24 times th~ African wage, while in other sectors the ratio was typically closer to the 7.3 observed in manufacturing, but was as low as 3.5 in financial services. Differentials tended to be largest in sector with the lowest skill content and were relatively stable within given sectors over time. To get an idea of to what extent the ovc~all wage differential was due to the inferior skill level of the African labor force, the European/African wage ratio is presented for a number of occupational groups in Table 6. The broad picture is that, excluding financial services, the European/Afri.can wage ratio is roughly around one half of the overall ratio and much less than this in agriculture. Of course, even within the defined groups, African wages may have been less than those of European because of disproportionate concentration of Africans within more lowly rated sub- occupations and because Africans were generally less experienced. The ratios thus overstate any purely discriminatory wage differentials that may have existed. A rough idea of the extent of pure wage discrimination may be obtained if one assumes that wages in public administration in 1981 were set entirely according to grade and experience, and by then comparing differentials in public administration with those in other sectors. This method is inaccurate insofar as skill distributions within observed occupation differ across sectors. In general, however, differentials in other sectors _, 15 - are rather similar to those in public administration with the noticeable exception of skilled and semi-skilled production workers in agriculture. It seems therefore, that African/non-African wage differentials at independence were mainly due to the immense skill disadvantage of African workers (much of this being due to discrimination with respect to access to training and education), as opposed to discriminatory wage differentials existing between identical workers. 4. The Non-Formal Sector Most of Zimbabwe's labor force is now outside the formal wage sector: either on communal lands, in the non-agricultural informal sector or simply unemployed. Communal farming is still by far the largest of these alternatives and employs around 1.4 persons or 40% of the total labor force at the current time, while non-formal activities taken as a whole employ at least 50%. Not only are non-formal activities a major source of labor income, but they constitute the prime source of unskilled African labor supply to the formal sector and given limits to the rate of growth of the latter, will have to act as a major source of labor absorption in the future if rapid growth in unemployment is to be avoided. (i) Communal Areas The communal areas are a combination of the Reserves of Southern Rhodesia, that were set aside for exclusive Africa occupation between 1890 and 1923, and further lands termed Special Native Areas that have been similarly allocated since then. At Independence, it was estim~ted that 700,000 peasant households farmed 16.3 million hectares of communal land. The figure now is probably much closer to 800,000. The main food crops grown are maize, - 16 - groundnuts, cotton and millets, while extensive cattle and goat rearing is also practised. Although cultivation techniques utilize hybrid seeds and modern inputs such as fertilizers and pesticides, yields per hectare are generally much lower than in commercial areas. In 1980, gross output per hectare on commercial farms was roughly four times that of communal areas. However, in value added terms, the ratio was only 2.4, thus reflecting greater intermediate input use in commercial agriculture. Communal farms are also generally more labor intensive in the sense that labor input use per hectare is generally higher. Although techniques have been improving in communal lands in recent years, there are serious problems of soil erosion in cultivated areas and of overgrazing in pastoral are~s. Communal areas have always been a major source of labor supply to the formal sector. This takes three forms: (i) seasonal migration to commercial farms during the cotton picking season; (ii) local supply as permanent workers to commercial farms from adjoining areas, and; (iii) longer term circular migration to urban and mining areas. Of these (iii) is clearly the most important and this is shown by the high proportions of households in communal areas with long term absentee members. Out of twelve regional studies carried out between 1977 and 1982, nine estimated that over 50% of communal households had one or more members currently absent and earning wages, while another two surveys put this figure at around 35% .. ~/ A more recent survey carried out in 1983 by Bonnevie~/ in a ward of Morewa-Kabatana 21 Evidence surveyed in Zimbabwe Agricultural Sector Study. Annex II, World Bank. December 1983. 3/ Henriette Bonnevie. The Effects of Migration of Heads of Households. Mimeo. 1987. - 17 - I • distr~ct in Mashona1and East found that 47% of village households were headed by a migrant. Equally significantly, 89% of all non-migrant household heads had worked as migrants in the past. This would suggest that almost all households experience out-migration at the same time or other. The importance of male migration is further reflected in disproportionate number of female communal farmers - 56.3% according to the 1982 Census. This proportion is unusually high by international standards. The ties between a migrant and his communal household are very strong. ' Migrants usually visit their families quite often and remit substantial sums home. The Zimbabwe National Household Survey (1983/4)~/ estimated that over six major provinces, the proportion of communal households receiving remittances varied between a low o£'31% in Mashonaland Central to a high of 48% in Mashonaland West. There seems ~o be some question, however, as to the importance of remittances relative to household incomes. Bonnevie's 1983 survey put remittances at 52% of household income in the rather limited areas sampled. In contrast the World Bank (1983) study reported that remi- ttances accounted for only 11% of household income and this estimate is consistent with unpublished results of other ongoing research, although amongst those households that receive remittances this proportion is believed to be much higher. It should be noted, however, that these figures understate the contribution that migration makes to household income as a whole, in that remittances are likely to be less than the net gain accruing to migration. Agriculture is not the only activity within communal areas. 4 / Zimbabwe National Household Survey Capability Programme. reports on Demographic Socio-Economic Survey, 1983/4. No.'s 1 to 6. c.s.o., Harare, 1984. - 18 - Manufacturing activiities include beer brewing (in~luding chibuku), handicrafts, furniture-making and knitwear, while there is also a growing l wholesale and retail trade network. Nonetheless, data from the Jackson survey indicate that non-agricultural activities only accounted for around 12% of total income. It is also clear that household$ in communal areas are often engaged in a number of virtually simultaneous activities. The National Household Surveys tried to clagsify persons according to whether they had jobs, whether they were working or not in the previous week, whether or not they were housewives, etc. It is obvious from the huge disparity in results from the various surveys, that many individuals fell under more than one heading, so that, for example, a person could have a job, be looking for work, be a housewife and attending an educational institution all at once. The overlap between school attendance and working or seeking work was particularly obvious and this reflects a high proportion of young people among the economically active. The internal labor market works in a familiar fashion. Essentially households with low endowment ratios of livestock, equipment and land relative to labor tend to supply themselves to those with higher endowment ratios. Ownership of livestock and equipment is very unequal in communal areas. For example, the National Household Surveys found that the proportions of households without cattle varied between 39% and 55% over the provinces examined, while only 2 or 3% had equipment such as a tractor, a truck, a car or a water pump. In spite of this, off-farm wage employment is not a major feature of communal areas, although labor exchange arrangements across households are believed to be commonplace. Access to land is, moreover, more evenly distributed than asset ownership with around 80% of household farming - 19 - two or more acres. Relative equality of land holdings would tend to limit the need for an internal labor ma~ket. The general picture is that migration seems to have increased in relative importance over time with a possible fall-of£ in more recent years. There may have been two opposing forces in operation. One would be that increasing population pressure has made the returns to labor in communal areas less attractive relative to labor incomes elsewhere. There are no data on communal wages, but the rather fragmentary evidence on real per capita incomes does not suggest any clear trend over time. The other force which would dampen migration is the reduction in urban job-finding prospects arising from the stagnation of formal wage employment over the past decade. Bonnevie, for example, draws upon earlier surveys of Musami ward in Murewa-Kubatana district which indicate that the proportions of men aged 15 to 55 years absent due to employment elsewhere went from 24.1% in 1948, to 46.4% in 1958 and to 67.4% by 1968. Migration may, however, have decreased in more recent years, as her om1 study found only 31% of single men over 17 years of age absent. (ii) The Urban Informal Sector .I Much recent analysis of labor markets in developing countries has focused upon the duality in urban area between a high wage restricted entry formal sector on the one hand and a market wage, free entry informal sector on the other. Prospective migrants from rural areas are perceived as viewing the informal sector as an alternative source of income while searching or queuing for a formal sector job. Higher formal sector wages or worsening rural incomes would then have the joint effect of both increasing labor supply to - 20 - the formal sector and expanding the numbers in the informal sector at lower wages. In such analyses, the informal sector is thus seen as an alternative to open unemployment for those who have not as yet found formal sector jobs. Historically, it is clear that the informal sector did not play this role in the Zimbabwe labor markets for two reasons: legislative intervention regarding urban settlement and lack of supportive intervention regarding formal sector wages. Before independence, restrictions on urban residence prevented Africans from working in town unless they had both legal residence and proof of a job, the more recent restric't.ing statutes being the African (Registration and Identification) Amendment Act and the Vagrancy Amendment Act, both passed in 1973. The informal sector could not then act as a refuge for migrant job-seekers, while the growth of open urban unemployment was restricted by the same regulations. The growth of urban areas therefore depended largely on non-agri'cultural formal sector growth, while informal activities were relatively more common in small African touwnships in communal areas. The other important feature in Zimbabwe was that prior to independence, there was no direct attempt by government to support African formal sector wages. This lack o£ institutional wage support would have limited the amount of excess labor supply to the formal sector and the role that the informal sector would have played may have played in their absorption. The Informal Sector Survey1/ of 1983 illustrated the following key features regarding the informal sector in Zimbabwe: (i) informal sector ~/ N. P. Moyo, R. J. Davies, G.C.Z. Mhone, L. Pakkiri, The Informal Sector in Zimbabwe, Dept. of Economics. University of Zimbabwe. February 1984. - 21 - participants were predominately self-employed and 90% had ownership rights; (ii) informal activities were more heavily concentrated in manufacturing and repair activities (70%) than in wholesale and retail trade; (iii) most informal sector participants appeared to have entered the sector after losing a formal sector job - only 16% gave rural work or rural unemployment as their previous occupation, while only 13% had been previous unemployed in an urban area; (iv) the average wage of those other than the self-employed was Z$71.5 per month which was almost identical to the contemporary wage of domestic servants. The median wage was, however, substantially lower.. A more recent study by Horn ~/ found that net incomes amdng market vendors were, after adjusting for hours worked, about half the minimum industrial wage. The Informal Sector Study seemed to broadly indicate that in Zimbabwe, not only does the informal sector have a different activity mix than in other countries, with less emphasis on trading; but that, as a result of past and present ~egulations, the sector is more concentrated in fixed premises. The sector was not heavily composed of aspirants to formal sector jobs with the possible exception of a minority of wage earners. Most participants in the survey indicated that they would not accept formal sector employment at Z$300 a month or less. The nature of the urban inf~rmal sector may nevertheless have been changing over the last four years given a relaxation in some of the earlier urban residency restrictions. It should be pointed out, however, that some constraints continue to exist as the Government has generally blocked the !1 Nancy E. Horn. The Informal Fruit and Vegetable Market in Greater Harare. Working Paper 4186. Dept. of Land Management. University of Zimbabwe. April 1986. - 22 - establishment of shanty towns near urban areas, and continues to enforce the vagrancy laws. Current indications are that ~nformal activities are expanding as they absorb rural migrants and are probably becoming a closer substitute for open unemployment. This process is taking place, however, at the cost of increased overcrowding in permanent dwellings and is liable therefore to natural limitations. .· 5. Labor Market Interventions · (a) Backgt·ound There are two broad labor market areas in which the Government has adopted an interventionist approach since independence: the determination of wages and the provision of worker job security. Essentially the Government is concerned about the existence of wide wage differentials in the formal sector I and low standards of living among those at the bottom of the ~iage ladder. Its objectives in introducing an incomes policy were to: (i) nat·row income differentials; (ii) raise the income levels and standards of living of those at the bottom; and (iii) contain inflationary pressures. The incomes policy has operated through two main instruments: I the introduction and subsequent revision of minimum wages; and ~i;;:-act control or limits upon wage increases according to income level. Initially the main policy objective was that of greater equity, although in recent years the other objective of restricting inflation has also been given explicit emphasis. Fears that minimum wage policy would lead to employment reduction led the government to restrict employers' rights to retrench workers in a series of measures enacted from 1980 onwards. Job security regulations thus initially arose as an antidote to the expected effects of wages policy. - 23 - {b) Minimum Wages The evolution of minimum wage rates over time is shown in both nominal and real terms in Table 7. Essentially the minimum wage policy was at its most aggressive in all sectors during the early part of 1982 but rates were alLowed to fall sharply at the revisions of September 1983. Subsequent revisions have pegged real minimum rates in agriculture and domestic ser~~ce at a level still substantially above that when minimum rates were introduced I in July 1980, while the real industrial minimum rate has been kept near to its • introductory rate. The path taken by real minimum wages is very different from that originally envisaged by the Riddell Commission in their report of Jl~ne 1981 in 'which the real industrial minimum was recommended to rise i.t, gradual stages to Z$115 in 1980 prices by Jult 1984. It appears that the Riddell Commission's recommendations were followed closely up to and including January 1982, as the Commission had set target rates at 1980 prices of Z$85 and Z$91 for January of 1981 and 1982 respectively, which compa.res closely with the actual rates of Z$83 and Z$90. Clearly the policy was thrown off course by the economic crisis in Zimbabwe which deepened during late 1982. The initial aim of the policy l;vas to raise minimum wages to provide an improved living standard for low paid workers as calculated from the poverty datum line (PDL). Apart from purely humanitarian considerations, the Commission believed that higher wages could actually he in th~ inter~sts of employees given increased productivity, reduced absenteeism, etc. Why such higher efficiency wages would nbt naturally arise in the unregulated sectors in the absence of such a policy remained, however, unclear, although it is I possible that Industrial Boards may have previously set unskilled African wages according to implicit market clearing criteria rather than acco~ding to efficiency wage considerations. - 24 - The negative effects of an effective minimum wage policy are well knoW1.1'; (i) reduced employment within individual production sectors and ccirrespondingly increased surplus labor and unemployment; (ii) a potential diversion of production from t'abor intensive to capital intensive sectors, from sectors facing price elastic product demand conditions to sectors facing more inelastic conditions and, given price controls, from rigidly administered sectors to sectors with cost-based price regulation such as manufacturing; (iii) a disproportionate fall in employment among young persons. For minimum wage policy to have induced any of the above effects, it would first have had to inflpence wages actually paid. This is further analyzed below in the general context of government wages policy. However, young, inexperienced worker constitute the group most likely to be affected by minimum wage legislation. This is particularly worrying in the case of Zimbabwe as open unemployment is most heavily concentrated among young, relatively less educated persons. Indeed, it may be no coincidence that the disproportionate incidence of unemployment in this group was observed at a time (August 1982) shortly after real minimum wages were at their peak. More direct evidence 1s provided by the declining number of indentured apprenticeships offered by private industry. Clearly an effective minimum wage makes it less attractive for em,toyers to provide training to young worke~s as there is less room for the worker to pay some of the training costs via lower wages, and it is less worthwhile for the employer to offer a rising wage scale to workers to indq.ce a lower quit rate. Although the number of African apprentices indentured has risen from 770 in 1981 to 1009 in 1985, the total number indentured has fallen from 2044 to 1197 over the same period. Thus although the racial imbalance in apprentice intake has been corrected - 25 - since independence, it may well be that during a period when the government was trying to expand the stock of trained manpower, its own labor market policies may have been a significant obstacle to achieving this objective. (c) Policy Towards Wage Differentials Following the publication of the RiddelL r~port, the Government introduced a graduated scale of permitted rates of wage increase as of January 1, 1982, which allowed larger percentage pay raises for those towards the bottom of the wage ladder. Any increase granted to a given category of employees was also to be granted to those on lower wages in the ratio of the stipulated maximum increases. The maximum increases permitted ranged from 23.5% for those earning less than Z$100 per month down to 1% per annum for those earning Z$1666 per month and above. A wage freeze was then in force until September 1983. Further adjustments included a Z$10 per month award to those earning Z$300 or less per month in 1983, following the removal of food subsidies in the same year; an.d a further discretionary award of up to 10% for those at the bottom of the scale in July 1984. The wage and salary review of 1985 introduced a new set of permitted maximum increases ranging from 15% at the bottom to 2% at the top. A more rigid policy phase has been applied during 1986/7 in which stipulated increases have been imposed. The anti- inflationary aspect of incomes policy has been given clear emphasis recently, with average pay raises set equal to half the expected rate of inflation. The Government can of course alter this formula by suitably changing its inflation II forecast. Exceptions are allowed at individual firm level, but block applications are discouraged. Zimbabwe is expected to adopt a formal incomes policy with a~ Q.ngoing incomes board in the near future. Table 8 gives some roughly comparable information on monthly wages in - 26 - October of 1982 and 1985 with corresponding annual average wage increases for some selected occupations in the private sector and·in public adminiutration. Although there is some indication of mild salary compression in the public servic~s, this does not appear to have been the case in the private sector. Given that the consumer price index rose at an average rate of 16.1% per year over these years, public service salaries generally declined in real terms, while of the occupations illustrated, only semi-skilled workers, technicians and, rather surprisingly, inexperienced clerks enjoyed real wage advancement. There is no doubt that collective bargaining agreements and the recommendations of Industrial Boards as accepted by the Government did impose wage compression by broad grades of workers. The conclusion must be therefore that in the private sector, firms have granted less than the maximum increases to non-scarce groups of workers such as the unskilled and copy typists, while regrading workers in scarce supply so as to avoid the regulations. It may well be therefore, that although the policy may not have done very well in achieving its aims of greater wage equity, it also did not contribute greatly to skill sho~tages in the private sector. Here the evidence is rather thin. However, the vacancy rates as calculated in the 1985 Annual Survey of Occupations are only near 10% or above in a few high level occupations sue~ as archltects, sales managers and veterinary workers. All of these are occupations within which the supply of skilled persons would have fallen·due to white emigration. The great majority of private sector occupation do not seem to be in a shortage situation as indicated by high vacancy rates. The more rigid policy stance taken in the last year or so is, however, believed to have exercised a greater influence upon private sector wage differentials, but the effects of this are not yet apparent. - 27 - The effects of recent wages policy upon the public services may, however, have been more serious. Private/public pay differentials now seem to exist from typists upwards and to widen as one moves up the skill ladder. There are two worrying implications of this. One is that the quality of the public service workforce may be impaired. Here warning might be served by the poor state of the civil services in other African countries that have already trocden further down the path of real public service pay reductions. Fortunately, Zimbabwe is still a long way from this position. The other problem is that shortages of certain key skills are sure to develop. This is already clearly evident with respect to some occupations such as computer programmers, and was reflected in significant 19~5 vacancy rates among legislative officers (13%) and government executive off1cers (10%).'• (d) Wage and Employment Effects of Incomes Policies To examine the z:tet effects of different wage policies, average wage . .. rates for the years 1980 to 1985 were compared with those as predicted from earlier real wage ~ehavior during the years 1955 to 1979. The impact of minimum wages should be seen most clearly in sectors with the highest unskilled labor content during the years 1981 and 1982 when minimum rates were at their relative peak. The effects of wage restraint and salary compression should, however, show up in later years, particularly in the more highly skilled sectors. This basic pattern is confirmed rather strongly in the results given in Table 9. In almost all sectors, wages were highest relative to predicted levels in 1981 and 1982, but fell from 1983 onwards. In sectors with the lowest skill content such as agriculture and domestic service, the minimum wage effect was particularly strong in 1982 and the minimum rate may still be I - 28 - effective in agriculture at the present time. The fact that the wages of domestic servants were below predicted levels in 1984 and 1985 may indicate that wage restraint is now the controlling factor in this sector. In contrast, as one would expect, the minimum wage appear to have no discernible effect upon the average wage in the high level manpower intensive financial services sector. Here salary compression and pay restraint have dominated \ ~ from 1982 onwards. In other sectors, one observes a mixture of policy effects. Minimum . been negligible even in 1982. wage effects in construction appear to have \ \1 ~ However, average wage rates in con·struction are highly pro-cyclical and would normally have risen relative to those other sectors during the 1981/2 boom, thus reducing the potential impact of an all-industry minimum wage rate. In mining, manufacturing, wholesale and retail, and transport and communication, minimum wage effects seem to have dominated through 1982 while other policies have become more important since. The net effect of the two sets of policies was to increase the average formal sector wage by 17% in 1982, although at present government policies have a wage restraining effect. Some indication o£ employment effects that minimum wage legislation might have had, can be inferred from some of the results presented in Table 10. In agriculture, the elasticity of employment with respect to the real product wage is estirnate.G at -0.44. Assuming given output and intermediate input prices in this sector would then suggest that a 40% increase in agricultural wages due to minimum wage legislation would in turn lead to an 18% £all in agricultural employment. The results also suggest that this reduction would be spread over several years. This result is broadly consistent with the actual behavior of agricultural employment in that the - 29 - number of workers employed in the sector fell by 20% between 1980 and 1983. In other sectors there seems to be little reason to believe that minimum wage legislation had much overall effect either, because there is no reason to believe that average wages were increased, or because 5 as in the case of manufacturing, the wage elasticity was near zero. In the formal sector as a whole, the maximum wage raising effect of 17% only translates into an employment reduction of around 4%, given a real product wage elasticity of - 0.46 and assuming that one half of nominal wage increases would be passed on in higher product prices. (e) Job Security Regulation From 1980 onwards, the Government has introduced regulations to restrict the laying off and firing of workers. The motivation here has been the preservation of as high an employment level as possible, particularly in view of the potential employment reducing consequences of minimum wages. The key questions are therefore whether or not these restrictions are likely to meet this objective and whether other undesirable side effects are likely to be incurred. In general there are three aspects of job security regulation: {i) the period of notice required; (ii) the formula by which compensation for non-disciplinary dismissal is determined; and (iii) government permission requirements. The essential feature of the regulations in Zimbabwe is that they revolve almost entirely around the need to obtain government approval. The period of notice is between one day and one month depending on the pay interval and the length of service, while the only legislated compensation is pay for the period of notice for immediate non-disciplinary dismissal. By international standards these last conditions are hardly severe. For example, - 30 - in Sudan, legal compensation is six months pay plus one month's paid notice; while in India, compensation is fifteen days pay for each year of service. It . should be noted, however, that more generous compensation and longer periods of notice may be recommended by Employment Boards or arise from collective bargaining in Employment Councils. The requirement that employers require government permission prioi to laying-off or retrenching employees first appeared under the Employment Act of 1980 and has since been clarified and extended in three further statutory instruments which appeared in 1981, 1982 and 1985. The essential modifications introduced in these later instruments have largely served to list acceptable reasons for disciplinary dismissal, to define penalties fdr unapproved, wrongful dismissals - currently reinstatement with back pay for the employee and a fine not exceeding Z$2000 and/or up to twelve monthss imprisonment for the employer; and to regularly reaffirm the permission requirement. The regulations apply to every undertaking, industry, trade or occupation except officers of parliament, unpaid work for a charitable institution and university students undertaking required employment as part of their curricula. Fixed term contract workers are also fully covered prior to the expiration of their contracts. Permission is not, however, required if both parties agree in writing to the termination of employment. Applications to lay off, retrench and to put workers on short--term working are dealt with by the Retrenchments Committee in the Ministry of L~bor~ After hearing evidence from the parties concerned including worker or trade union representatives, the Committee makes its recommendations to the Minister. The indications are that applications for retrenchme~ts are rarely granted, although the Committee may recommend interim measures such as short- .. - 31.- term working and temporary unpaid leave. The procedure can therefore be a lengthy one and some applications may be considered over a period as long as one year. Apart from any short-term employment preservation effects, the justification for government intervention would be that in the absence of such regulations, .;ontracts voluntarily struck between workers and employers would imply excessive employment instability. Essentially the argument is that in an optimal contract, risk averse workers would voluntarily accept lower wages in return for guaranteed employment in a time of recession. In practice, however, workers may not be willing to accept such contracts for fear that some employees wilt break their word when downturns actually occur. Imposition of additional job security by government could then be seen as a correction for a labor market imperfection arising from non-enforceability of contracts. Two of the preconditions of this argument do not, however, apply in Zimbabwe. Labour contracts are enforceable through appeals to the relevant Employment Board or Council while there is also a well developed legal system. Concern about non-enforceability could then be averted through some reinforcement of these institutions. The other point is that the system of administered wages and, in particular, that of minimum wages, may prevent employers from offsetting ~he expected costs of retaining an excessive workforce in times of recession by offering lower average wages over the cycle as a whole. A more aggressive minimum wage policy would .. thus increase the costs imposed on employers by job security legislation. The effects of job security regulations are twofold: to raise the implicit cost_o£.labor input and tq retard adjustments in the labor force. Diminished labor productivity is likely to result from inability to dismiss - 32 - less productive workers and from increased labor shirking. In addition, however, firms face the increased costs of either paying excess labor during periods of contraction in labor demand, or of paying workers to voluntarily quit or retire. This last practice seems to be commonplace throughout the formal wage sector with typical compensation running at around eight months pay. Employers will, of course, try to substitute temporary for permanent workers. However, this process is likely to be limited for two reasons. Firstly, casual labor has to be paid at twice the legal minimum rate and cannot be used for more than six weeks in any three month period, while contractors are subject to the same job security legislation as everyone else and their use seems therefore to be rather limited. Secondly, temporary labor is generally likely to be of inferior quality to permanent staff, which, given the same or higher wage rates, increases the current cost per unit labor input. The only direct evidence available on this is drawn from commercial agriculture, within which 73% of all employees held permanent status in 1980 as compared with 70% in 1984. . The combination of minimum wages and imposed job security legislation appears then to have had only a limited substitution effect of permanent for temporary employees in this sector. The longer term effects of a continued imposition of the government permission requirement are likely to'be that: given higher expected employment costs, firms will adopt more capital intensive techniques; they will adjust hours worked subject to cost and institutional constraints over the cycle rather than employment; and they will become less responsive to improved market conditions such as an increase in export demand. Two adverse macroeconomic consequences would then follow. The overall growth in labor - 33 - demand will be more limited than otherwise and this will be especially serious in view of Zimbabwe's current and projected unemployment problem; while on the product market side, the ability of the formal sector to structually adjust to internal and external conditions will be impeded. . .. - 34 - .. Table 1 Employment by Broad Sector: 1969 and 1982 . ,, 1969 1982 % % Formal ·wage employment 44.6 47.2 of which: Commercial Agriculture 16.1 12.4 Non-Agriculture 27.8 34.8 Small Scale Farms 0.7 0.7 Communal Farmers 53.4 46.9 Informal 2.1 5.9 100 100 Total No. (OOO's) 1604,7 2215.9 Sources: Census of Population 1969. c.s.o., Harare 1976. Main Demographic Features of the Population of Zimbabwe: An Advance Report Based on a Ten Percent Sample. c.s.o., Harare 1985. Quarterly Digest of Statistics, c.s.o., Harare June 1986. Note: (i} Employment here refers only to persons 15 years or above. In 1969, younger workers accounted for 4.3% of the total labor force. This figure should have fallen considerably during the 1980s. (ii) Small scale farms are African owned. These are usually treated separately from commercial agriculture. - 35 - Table 2 Unemployment Rates by Age, Sex an Education in August 1982 All No P·rimary Secondary Education Education Education Males 15-19 12.9 23.4 12.9 9.5 20-24 15.9 18.4 17.5 10.7 Total 10.9 11.6 11.7 7.0 Females 15-19 17.0 22.7 16.9 13.2 20.24 15.8 13.7 17.0 13.4 Total 10.7 8.8 11.6 10.5 Persons 15-19 15.0 22.9 14.9 11.1 20-24 15.8 15.7 17.3 11.6 Total 10.8 10.3 11.6 8.1 Total No. 268,100 54,640 183,340 30,120 Source: Main Demographic Features of the Population of Zimbabwe: An Advance Report Based on a Ten Percent Sample. c.s.o., Harare, June 1985. .... 36 - Table 3 Distribution and Gro~1th Rates of Wage Employment by Sector 1985 1955-65 1965-75 1975-80 1980-85 % % % % Agriculture, Forestry, etc. 0.26 0.6 2.1 -2.1 -3.3 Minin$ and Quarrying 0 .• 05 -·.1,!99 5.7 2.9 -4.0 Manufaeturing 0 •. 16 1.4 6e6 0.4 1.4 Electricity, Gas, etc. 0.01 -0.1 3.4 -0.6 3.0 Construction 0.04 -5.7 8.0 -7.3 1.6 Wholesale and Retail 0.07 1.6 2.4 -1.9 2.3 Financial Services 0.01 7.1 4.3 0.7 4.3 Transport and Communication 0.05 1.6 2.5 0 .. 1 5.1 Public Administration 0.09 3.6 4.9 7.5 5 .. 1 Education 0.08 5.9 2.• 1 3.0 15.1 Health 0.02 6.3 3.9 2.4 5.3 Domestic Servants 0.09 I 1.8 2.7 -2.8 -1.9 Other Services 0.06 4.4 5.3 0.4 6.0 Total 1.0 1.3 3.4 -0.8 1.0 Sources: Quarterly Digest o£ Statistics: June 1986 Monthly Dige~t of Statistics, June 1965 and 1965 Data for 1985 supplied by c.s.o., Harare. Note: The armed services are shown under public administration. ·- 37 - Table 4 Percentage Skill Distribution by Sector (1981} Professional Skilled Semi-Skilled Unskilled Agriculture, Forestry~ etc. 1 5 13 81. Mining & Quarrying 3 12 26 59 Manufacturing 4 13 26 57 Electricity 5 15 31 49 Construction 2 17 11 71 Wholesale & Retail 7 16 31 45 Financial Services 17 16 39 38 Transport & Communication 7 16 39 38 Public Administration 12 11 25 52 Education 53 33 4 10 Health 34 6 22 38 Other Services Total 8 12 20 60 Source: National Manpower Survey 1981. Ministry of Manpower, ~1anning and Development. Notes: (1) Domestic Servants ~re omitted (2) A slightly different industrial classification was used from that given in the annual employment survey. 'Other services' is incomparable with Table 4r (3) Professional - occupations for which a tertiary level qualification or equivalent e~pe~ience is required. Skilled - as designated under National Industrial Council agreements and includes all 'artisans, skilled manual workers, etc. (Clerical and sales workers with diplomas or doing jobs requiring secondary qualifications are also included. Semi-skilled - also as designated under Nation~l Industrial Counc~l ..~greement.s and includes all assistants and a wide range of oper4tors.who have accumulated experience on the job~ Unskilled - no experience or training required. - 38 - trable 5 Average Annual Percentage Increase in Real Consumption Wage by Sector . 1955-65 1965-75 1975-78 1980-85 Agriculture African 0.3 0.7 Non-African 1.9 1.5 Total 0 .. 8 0.9 8.7 Mining African 2.4 1~9 1.2 Nonoi'