From migrants to proletarians : employment experience, mobility and wages in Tanzania

World Bank Reprint Series: Number 250 J. B. Knight and R. H. Sabot From Migrants to Proletarians Employment Experience, Mobility, and Wages in Tanzania Reprinted with permission from Oxford Bulletin of Economics and Statistics, vol. 44, no. 3 (August 1982), pp. 199-226.  World Bank Reprints No. 215. Michael Cernea, "Modernization and Development Potential of Traditional Grass Roots Peasant Organizations," Directions of Change: Modernization Theory, Research, and Realities No. 216. Avishay Braverman and T. N. Srinivasan, "Credit and Sharecropping in Agrarian Societies," Journal of Development Economics No. 217. Carl J. Dahlman and Larry E. Westphal, "The Meaning of Technologi- cal Mastery in Relation to Transfer of Technology," Annals of the American Academy of Political and Social Science No. 218. Marcelo Selowsky, "Nutrition, Health, and Education: The Economic Significance of Complementarities at Early Age," Journal of Develop- ment Economics No. 219. Trent Bertrand and Lyn Squire, "The Relevance of the Dual Economy Model: A Case Study of Thailand," Oxford Economic Papers No. 220. Gershon Feder and Knud Ross, "Risk Assessments and Risk Premiums in the Eurodollar Market," Journal of Finance No. 221. Bela Balassa, "Policy Responses to External Shocks in Selected Latin American Countries," Quarterly Review of Economics and Businless No. 222. Choong Yong Ahn, Inderjit Singh, and Lyn Squire, "A Model of an Agricultural Household in a Multi-crop Economy: The Case of Korea," Review of Economics and Statistics No. 223. Emmanuel JimeE-rz, "The Economics of Self-help Housing: Theory and Some Evidence from a Developing Country," Journal of Urban Economics No. 224. Thawat Watanatada and Clell G. Harral, "Determination of Economi- cally Balanced Highway Expenditure Programs under Budget Con- straints: A Practical Approach," Transport Research for Social and Economic Progress No. 225. George Psacharopoulos, "The Economics of Higher Education in Developing Countries," Comparative Education Review No. 226. Katrine Anderson Saito and Delano P. Villanueva, "Transaction Costs of Credit to the Small-scale Sector in the Philippines," Economic Development and Cultural Change No. 227. Johannes F. Linn, "The Costs of Urbanization in Developing Coun- tries," Economic Development and Cultural Change No. 228. Guy P. Pfeffermann, "Latin America and the Caribbean: Economic Performance and Policies," Southwestern Review of Management and Economics No. 229. Avishay Braverman and Joseph E. Stiglitz, "Sharecropping and the Interlinking of Agrarian Markets," American Economic Review No. 230. Abdun Noor, "Managing Adult Literacy Training," Prospects No. 231. Bela Balassa, "Shifting Patterns of World Trade and Competition," Growth and Entrepreneurship: Opportunities and Challenges in a Changing World No. 232. Johannes Bisschop, Wilfred Candler, John H. Duloy, and Gerald T. O'Mara, "The Indus Basin Model: A Special Application of Two-Level Linear Programming," Mathematical Programming Study  FROM MIGRANTS TO PROLETARIANS: EMPLOYMENT EXPERIENCE, MOBILITY AND WAGES IN TANZANIA* J. B. Knight and R. H. Sahot I. INTRODUCTION It is characteristic of economic development that there emerges a stable and committed proletariat possessing the skills needed by the expand- ing modem sector of the economy.l Our purpose in this paper is to describe and explain the structure of incentives which fosters this process, illustrating from a survey of manufacturing employees in Tanzania. We wish to discover whether, how, and why employment experience is rewarded by employers, and what are the determinants and consequences of the extent of labour mobility. The wage labour force in Tanzania has indeed undergone a transition from short-term migrant to proletarian status. For the first half of the twentieth century migration for wage employment in Tanzania was circular and short-term. Absence from home was generally for less than a year and rarely for more than four years.2 By 1971, however, the situation in manufacturing was quite different. No less than 68 per cent of the employees in our manufacturing sample were migrants in the sense that they had arrived in the city at the age of 14 or over. How- ever, the estimated rate of labour turnover, i.e. the number of changes in employees in a given stock of jobs, was roughly 1 per cent per month, considerably lower than in US and British manufacturing.3 The estimated rate of mobility, i.e. the number of job changes by a given stock of employees, was 0.137 per year of employment; the rate of voluntary mobility was still lower, 0.081.4 These measures of the stability of the manufacturing labour force suggest that post-independ- ence Tanzania is closer to the Japanese model of lifetime commitment to the firm than the African colonial model of a labour force comprised almost entirely of temporary employees. * We are grateful to Joy de Beyer, Chris Gilbert and the Editors for comments on an earlier draft. I Accounts of this process occurring in African countries are to be found in Elkan (1960) for Uganda and Sabot (1979) for Tanzania. 2Sabot (1979), Ch- 7. 3Official statistics of turnover in 1971 indicate rates of about 4 and 2.5 per cent respectively. 4 See Table 6 for other indicators of stability and methods of estimation. 199  200 BULLETIN The increase in employment stability and corresponding increase in residential stability of Tanzania's urban labour force occurred in response to changes in economic conditions. In the colonial past the wage structure in agriculture, which provided most wage employment, and in much of the urban sector, did not offer incentives for workers to become experienced. The wage was competitively determined at the supply price of temporary migrants, and the earnings-experience profile was typically horizontal. No training period was required for entry into the great majority of jobs, nor did these jobs provide opportunities for learning-by-doing. For employers the benefits of having an experienced labour force were negligible: crude physical labour was enough. As long as the demand was for unskilled labour, the high turnover rates implied by the system of circular migration did not adversely affect profitability. By 1955, however, concern was expressed5 that a modern industrial sector would not prove economically viable if workers continued to be so uncommitted to their jobs that, soon after acquiring skills, they returned to their rural areas to be replaced by raw recruits. This con- cern spread as non-agricultural wage employment grew in importance and the number of jobs involving skill-acquisition increased, and it was a factor motivating both government and private wage policies. The new wage policies markedly increased incentives for workers to acquire employment experience. The substantial rise in urban wages relative to rural incomes, the result in part of the introduction of the minimum wage in 1959 and subsequent increases in its level, was perhaps the principal means by which the system of short-term circular migration was eroded.6 In the dozen years prior to the survey, non- agricultural real wages trebled whereas real income per capita in agricul- ture stagnated.7 Higher urban wages provided both the incentive and the opportunity for stabilization: they raised the opportunity cost of leaving wage employment for spells of tending the family shamba, and they made it possible for migrants to support their families in town. The rise in wages eventually generated an urban labour supply in excess of demand, so providing incumbents with a greater incentive to hold on to wage jobs. But there was another factor at work. The introduc- tion of wage structures which rewarded experience also increased the opportunity cost of leaving wage employment. The incentives for manufacturing workers to stay in employment were shown by our survey to be substantial. A worker with the mean length of time in wage employment (nine years) received on average 40 per cent more than an entrant possessing the same characteristics. Moreover, firms' wage structures encouraged employment stability not only within the 'In the Report of the East African Royal Commission. 6 Sabot (1979), Chs. 7-8. 7Sabot (1979), Tables 2.5, 2.8.  FROM MIGRANTS TO PROLETARIANS 201 urban areas but also within firms. Our focus in this paper is on the nature of these rewards for the accumulation of employment experience. We are able to explore the relationship between employment experi- ence and earnings, and the reasons for this relationship, in ways which represent an advance on existing work. Almost all the literature which examines the relation between skill-formation and earnings uses a variable of the form (A-E-6) as a proxy for employment experience, where A is the age of a worker and E hiis years of education. Our survey provides information not only on the actual number of years of wage employment but also on length of employment in the current job and in previous jobs; and we have information on the extent of and motives for inter-firm mobility of our sample. This evidence enables us to examine and compare hypotheses about the structure of earnings which are drawn from the 'human capital' and 'internal labour market' theoretical frameworks. The survey from which our evidence is drawn was conducted by one of the authors in 1971. It covered some 1,000 employees in the manu- facturing sector of Dar es Salaam. The sample was based on a random selection of 47 firms, stratified by size of firm, and random sampling of employees within each firm. Detailed interviews were conducted with each of the respondents, and employment histories pieced together. Section II of the paper provides a conventional account of the relation between aggregate employment experience and earnings. By contrast, Section III disaggregates employment experience to examine the distinction between marketable and non-marketable human capital, i.e. general and firm-specific skills; it also compares human capital with non-'human capital', and external with internal labour market inter- pretations of the findings. Section IV presents some relevant results on the causes and consequences of inter-firm mobility. Section V concludes. II. MARKETABLE HUMAN CAPITAL The conventional treatment of employment experience in earnings function analysis assumes that it gives rise to general skills which, in competitive markets, possess a market value. Since stuch skill acquisition raises the marginal product of workers equally in many employments, competition among employers ensures that the wage rises to maintain equality with the marginal product. The employer cannot capture any of the return on the costs involved in training or in learning-by-doing, and is therefore unwilling to pay for these costs. During the period of skill-acquisition the wage received is the actual marginal product minus any training costs. The graph of wage against employment experience is likely initially to be upward sloping: low at first because training costs are deducted from the wage but rising as the product of the worker increases and as training diminishes. If skill acquisition ceases, the wage  202 BULLETIN curve reaches a peak, and if human capital depreciates with time, the wage then begins to fall. Hence the likelihood of a concave wage- experience profile in the case of general skill acquisition. Table 1 presents the results of an earnings function analysis in which years in wage employment (L2) is included as a continuous variable to represent on-the-job skill acquisition. In order to test for non-linearity in the relationship between earnings and experience, the squared term (L2)' or the dummy variables L21 (1-3.9 years of employment experi- ence), L22 (4-7.9 years), L23 (8-11.9 years) and L24 (12 or more years) are introduced. Employment experience is thus incorporated as L2, or as L2 and (L2)2, or as the four L2i dummy variables. Other personal characteristics are included in the equations8 but only the variables representirig skill acquisition in employment are reported. The depend- ent variablo is the natural logarithm of earnings (Z). The analysis relates to thle /28 African regular employees in the sample. Significant coefficients are obtained for all three forms of the experi- ence variable, and it is difficult to choose among them. When L2 alone is included, the coefficient is positive and very highly significant; it implies an increase of 3.1 per cent in earnings for each additional year of wage employment. The inclusion of (L2)2 as well as L2 hardly improves the predictive power of the equation. The value of the coeffi- cient on L2 is nearly doubled, to 0.052, and it remains significant at the 1 per cent level. The coefficient on (L2)2 is negative, as predicted, and significantly so. The coefficient implies a peaking of the earnings- experience profile after 29 years in wage eniployment;9 at the peak, experience adds 75 per cent to earnings. Fewer than 2 per cent of the sample would have gone beyond the peak. The use of experience dummy variables again produces significant positive coefficients (with less than four years of experience being the base dummy). The coeffi- cients rise consistently up to and including the highest experience category (12 or more years of experience). In sum, the prt. ums earned by more experienced workers are large and subject to only slightly diminishing returns.'0 Another result is in line with the marketable human capital explana- tion of the experience-earnings relation. Whether or not a worker has received formal training is included as a dummy variable among the independent variables. The coefficient on F2 in Table 1 indicates that 8 See the notes to Table 1. 'The peak is derived by differentiating the function a(L2) + b(L2)' and solving a + 2b(L2) = 0. '