Gold Digging Careers in Rural East Africa: Small-Scale Miners’ Livelihood Choices

World Development Vol. 38, No. 3, pp. 379–392, 2010� 2009 Elsevier Ltd. All rights reserved

0305-750X/$ - see front matter

www.elsevier.com/locate/worlddevdoi:10.1016/j.worlddev.2009.09.003

Gold Digging Careers in Rural East Africa: Small-Scale Miners’

Livelihood Choices

DEBORAH FAHY BRYCESONUniversity of Glasgow, UK

and

JESPER BOSSE JØNSSON *

University of Copenhagen, Denmark

Summary. — Rural livelihood studies over the past two decades have stressed directional movement away from smallholder agricultureand the diffuse, ad hoc, uncertain, and low-earning character of most rural non-agricultural income diversification. Based on a recentsurvey of small-scale mining in Tanzania, this article documents the higher risks, greater potential earnings, more elaborate divisionof labor, and career trajectory of miners. Tracing cohort entry groups indicates that those willing to withstand the hardships of movingfrom one gold strike to another and time commitment to a career considered dangerous and alienated from agrarian traditions of theTanzanian countryside may be materially rewarded for their efforts.� 2009 Elsevier Ltd. All rights reserved.

Key words — small-scale mining, careers, livelihoods, income diversification, labor specialization, Tanzania

* We are grateful to Rose Mwaipopo, John Wihallah, and the late Nda-

lahwa Faustin Madulu for their advice and comments during our research

in Tanzania’s gold fields and to Niels Fold and Per Kalvig, as well as

anonymous reviewers, for their comments on earlier drafts of this paper.Final revision accepted: July 22, 2009.

1. INTRODUCTION

Much has been written about income diversification in ruralAfrica over the past two decades. The decline of smallholderfarmers’ commodity prices and productive and marketinginfrastructure ushered in a protracted period of rural liveli-hood experimentation. The literature has debated what effectthe trend is having on levels of rural poverty and inequality(Ellis, 2000; Ellis & Freeman, 2005; Homewood, 2008; Mea-gher, 2001). Profound occupational restructuring through pro-cesses of ‘‘deagrarianization” and ‘‘depeasantization” isinsidiously altering the social, cultural, and political fabric ofthe African countryside (Bryceson, 1996, 2000; Bryceson & Ja-mal, 1997).

As increasing amounts of rural household income began tobe generated through non-agricultural activities, it became un-clear what work was replacing smallholder commercial agri-culture. Rural trade and service activities proliferated, but ascrutiny of individual and household income-earning revealeddiversified ad hoc work experimentation of a highly fluid nat-ure rather than regular, specialized work patterns. Hence thelivelihood literature of the last decade and a half has focusedprimarily on the assets and capital investments that ruralagents could deploy in their experimental livelihood activitiesrather than any targeted skill acquisition for defined careerpaths. Livelihoods were in flux and donor agencies saw theirrole as facilitating producers’ credit or infrastructure for muchneeded rural household income-earning in the short and med-ium term.

In the process of deagrarianization, a number of African na-tional economies have become increasingly reliant on mineralexports, and some of them have afforded small-scale minersconsiderable space in prospecting and excavation. Using casestudy material from Tanzania, the following article documentshow the skill and teamwork requirements of small-scalemining 1 have, over the course of the last three decades, pushed

379

beyond livelihood experimentation to coalesce as discernablecareer paths in mining. Hundreds of thousands of peopleacross the age spectrum are now earning a livelihood insmall-scale mining. The following discusses why and how thissector of the economy is attracting workers toward it, therebygenerating an alternative to peasant farming, despite publicperceptions of exploitative and insecure work conditions andsocial vice in the mining settlements.

The first section contrasts experimental livelihood diversifi-cation with features of mining that militate for the emergenceof small-scale mining careers. The rise of Tanzanian small-scale gold mining as a magnetic force in rural labor marketsis reviewed before turning to the methodology and back-ground of our comparative study of two Tanzanian gold min-ing settlements. The surveyed miners are grouped into careerentry age and mining entry year cohorts as well as hierarchicalmining function groups for tracing their careers and differenti-ated outcomes before turning to the distinction between careerand diversification strategies in our two case study miningsites. The conclusion appraises the career patterns of Tanza-nian small-scale miners at present and in the future.

2. LIVELIHOOD EXPERIMENTATION VERSUSCAREER TRAJECTORIES IN RURAL SUB-SAHARAN

AFRICA

The African rural livelihoods literature (Bryceson, 1999,2002a; Ellis, 2000; Ellis & Freeman, 2005; Francis, 2000;

380 WORLD DEVELOPMENT

Grawert, 1998) has extensively documented the diversified ar-ray of service and productive activities that country dwellershave sought as alternatives to commercial farming. Whilesome livelihood typologies have emerged, researchers havenot generally identified regularized activity sequences andoccupational patterns in their survey data and work histories(Bryceson, 2002b). Instead, the recorded work activities andtheir temporal ordering in individual work lives have beencharacterized by open-endedness, erratic change over time,and/or the simultaneous pursuit of multiple activities, relatedto the extremely competitive conditions that diversifyingexperimenters have faced.

The lack of agricultural investment in African countries andthe severe disincentive effects experienced by African small-holder farmers in global commodity markets in the wake ofthe 1970s oil crises pushed rural smallholder farming house-holds to reduce production of the agricultural export cropsthat African economies had been identified with for decades.Instead, they sought to construct alternative income-earningportfolios against considerable odds. Rural purchasing powerwas contracting in response to the shrinkage of agriculturalcommodity markets. Large numbers of rural dwellers, notablyyouth and women, were flocking to easy-entry service andproduct markets where start-up capital was not required. Gi-ven the magnitude of labor displacement from commercialagricultural production, the supply of goods and services inthese markets became glutted, driving prices down. Lack ofcapital and skills kept most income diversifiers restricted tothese markets.

Only a few, usually those with the advantage of strong ur-ban connections or ample land assets, had the capital to enterthe more lucrative fields of grain milling, motorized transportservices, and guesthouses. In the 21st century, many parts ofrural Africa evidenced increasingly polarized income profileswith large pools of income diversifiers, devoid of occupationaltrajectories, competing with one another, lowering their aver-age earnings, as opposed to small numbers of successful localinvestors with sufficient capital to go into high-yielding trans-port services and trade (Meagher & Mustapha, 1999).

Unlike rural Asia and Latin America, where formal wagedwork prevails, informal self-employment has been dominantin Africa. As opportunity costs of labor declined, employersincreasingly saw rural labor as cheap and disposable. Whereplantation formal wage labor had existed, it tended to becomecasualized (Konings, in press). Meanwhile, traditional laborapprenticeship arrangements disintegrated as employers be-came averse to investing in skills training (Meagher, in press).Most engaged in subsistence agricultural production, whichfunctioned as their fallback in the event of insufficient earningsfrom their non-agrarian activities.

Massive numbers of rural dwellers resorted to petty tradeand low capital entry services such as charcoal burning andbeer brewing. Traders were vulnerable to loss of their tradingcapital, particularly in oversupplied local perishable food mar-kets. It was in this context that mining became a magneticforce in mineral-rich countries, particularly those like Tanza-nia, where small-scale mining had been discouraged or ignoreduntil relatively recently (Hilson & Potter, 2005).

Beyond legal constraints, the extent and nature of small-scale mining in Africa depends on an array of factors (Hilson,2003). The richness of the mineral deposits, small-scale miners’technological capabilities, and not least the stage at which goldmining exploration, production, and export have advanced inthe country. Africa has a broad range of mineral-endowedcountries, from South Africa where gold mining has been pur-sued for over a century, to countries like Tanzania, where gold

mining as a small-scale mining activity has surged into a massactivity relatively recently through a series of gold strikes.

Generally, at the outset of being attracted to gold mining,rural dwellers experiment, as they would in any other non-agrarian income-earning activity, but there are notable differ-ences since gold mining is, first, perceived to have potentiallyfar higher earnings; second, it can lead its practitioners awayfrom the security of their home areas in a quest to strike goldwherever that may occur; and third, it necessitates networkingand mining skills that are generally acquired informally overtime. As a consequence, compared with other activities in ruralAfrica, gold mining tends to attract risk-taking maximizers.Furthermore, as miners progress from one mining site to an-other, those who have proven competency in mining andextensive social networks affording them timely informationon new strike sites and coalescing mining teams have a strongadvantage (Jønsson & Bryceson, 2009).

Taking the comparison of trading and mining further,whereas trading and rural service provisioning is a competitivescramble that marginalizes many, mining can become a steelierendurance test. Unlike rural trade in petty commodity and ser-vices where purchasing power is lacking, gold has a compara-tively better domestic market linked to the international goldmarket, which has been exceptionally buoyant in recent years.

Those attracted to small-scale mining generally enter with-out specialized mining skills. In effect, a miner’s first work siteis an apprenticeship where he earns as he learns. Gold miningis straightforward at the initial panning stage, but as goldavailability ‘‘sinks deeper,” technical skill is essential for find-ing gold. Pits usually reach 10–50 m in depth and cave-ins arenot uncommon. Social skills and esprit de corps are equally vi-tal. Miners have to be capable of hard work, trust, coopera-tion, and sobriety to successfully and safely find and extractthe gold.

Having compared the mining process to trade, we considerthe more marked differences with agricultural labor processes.First, it should be noted that small-scale mining generally co-exists and may rely heavily on the existence of smallholderagriculture as well as trade. The relative balance between min-ing, agriculture, and trade in a rural mining settlement is infact an indicator of the stage of small-scale mining underway,as will be discussed in Section 7.

Traditionally, African smallholder agriculture rested onfamily production units working land usually on the basis ofcustomary usufruct rights, in long-established local communi-ties where people had patrilineal or matrilineal ties to one an-other. Work tasks were generally allocated by age and genderwithin the confines of familial authority structures. Subsis-tence food production was given precedence. Cash croppingwas primarily within the purview of the senior male in the fam-ily.

Mine labor at strike sites 2 offers a stark contrast: mobileand multi-ethnic, with a continually high turnover of non-familial people working in close contact with one another.In view of the recent surge in mining and entry of large num-bers of mine laborers from agrarian and trading backgrounds,we need to ask how small-scale mining has coalesced so effec-tively, managing to achieve work efficiency and avoid mayhemin the production and processing of such a valuable commod-ity despite the deep rural poverty of the Tanzanian country-side.

The existing livelihood literature does not directly addressthis question.3 More revealing is the current literature on laborrestructuring and changing career patterns in the westernworld where analogous trends are occurring. Deindustrializa-tion has altered the traditional career structures of blue and

GOLD DIGGING CAREERS IN RURAL EAST AFRICA: SMALL-SCALE MINERS’ LIVELIHOOD CHOICES 381

white collar workers, casualization of labor has become perva-sive in many sectors, and gender opportunities in the workforce are being reconfigured. The interaction between the sec-toral change of deindustrialization, altered work opportuni-ties, and reconfigured career patterns, bears similarity to theprocess of deagrarianization, the surge in new mining oppor-tunities, and the evolution of mining career patterns in Tanza-nia’s small-scale mining sector.

There are innumerable definitions of careers. In generalterms, a career comprises an occupation taken up for a signif-icant period of one’s life with opportunities for progress. Thispertains to people who have worked in mining over a numberof years. Kanter (1989) identified conditions for the existenceof career trajectories including: first, long-term employment,allowing for the eventual reward of sacrificing behavior atthe beginning of one’s career and upward progression in ahierarchically ranked pyramid; second, sustained sectoralgrowth, allowing for the expansion of the pyramid; and third,a restricted number of competitive achievers with organiza-tional and productive roles who maintain coherence and en-force the legitimacy of the production pyramid bothinternally and externally.

The length of occupational engagement, the level of concen-tration on mining activities, and the willingness to move in re-sponse to the changing availability of minerals are the keyindicators of career commitment and progression, which dis-tinguish career miners from African rural income diversifiersmore generally. We define miners as people who are engagedin the active and enduring search for and extraction of miner-als. Those who have mined as diggers for more than 2 yearsand/or who proceed to another mining site can be consideredto have broken ranks with rural income diversifiers. This arti-cle traces the path of deepening mine work specialization andcareer development marked by the bifurcation of career-minded miners from diversifying part-time miners who stayin situ and pursue mining alongside other income-earningand subsistence activities.

3. OPPORTUNITY KNOCKS: SMALL-SCALE GOLDMINING IN TANZANIA

Despite Tanzania’s mineral wealth, mining was not widelypursued in the first two decades after the country’s indepen-dence in 1961 except in some localized areas. However, this be-gan to change in the 1980s. Since then, the country hasundergone comprehensive economic deregulation throughthe implementation of the Structural Adjustment Programmes(SAP) and economic liberalization during the 1990s. Miningand investment legislations were reformed, triggering growthin the small and large-scale mining sectors of the nationaleconomy (Butler, 2004; Phillips et al., 2001). The MiningAct of 1998 and the Mining Regulations of 1999 (Tanzania,United Republic, 1998, 1999) accord small-scale miners theopportunity to own Primary Mining Licenses (PMLs). 4 Con-sequently, small-scale mining sites have proliferated over thepast three decades leading to an estimated 300,000 small-scalegold miners and just as many extracting other minerals andprecious stones (Drechsler, 2001).

Small-scale mining constitutes a major employer and liveli-hood support for families throughout Tanzania. The impor-tance of the sector to local communities and the nationaleconomy is now officially acknowledged by the National Strat-egy for Growth and Reduction of Poverty, which states that‘‘small-scale mining is increasingly becoming dynamic as itprovides alternative economic opportunities to the rural com-

munities” (Tanzania, 2005, p. 7). At the same time, the sectoris associated with a number of immediate and long-term ad-verse effects (Phillips et al., 2001). Major gold strikes attractthousands of miners within a matter of weeks, turning smallvillages into burgeoning settlements characterized by provi-sional housing, overcrowding, poor sanitation, and limitedsecurity (Bryceson & Mwaipopo, 2009; Jønsson & Bryceson,2009). At the peak of a gold rush, excessive consumption ofconsumer goods, women, alcohol, and gambling are part ofeveryday life. The mining sites are of variable quality in termsof gold output and over time become increasingly costly anddifficult to work given the technological limitations small-scaleminers face as the gold is depleted and mining requires extrac-tion work at increasingly lower depths. Environmental degra-dation through mercury pollution, forest clearing, and lack ofland rehabilitation typically follows in the wake of goldstrikes. When miners move on, these problems are usually leftbehind. Closely associated with the life of a small-scale minerare health hazards such as injuries from accidents, exposure tosexually transmitted diseases, silicosis, and mercury poisoning(Fisher, 2007; Mwaipopo, Mutagwaba, Nyanga, & Fisher,2004; Phillips et al., 2001).

In general, the adversities of cost, time, and bureaucracyfaced by small-scale miners when attempting to formalize theiractivities through the acquisition of licenses exist in manycountries in sub-Saharan Africa (Hilson, Yakovleva, &Banchirigah, 2007; ILO, 1999). In Tanzania, while preciousmineral discoveries are often made by small-scale miners or lo-cal people, the restricted capacity of mining authorities totimely disseminate legislative information on how to acquirePMLs favors large-scale mining companies over small-scaleminers. The former secure licenses in mineral-rich areas beforethe majority of small-scale miners know of the opportunity.Thus, in effect, Tanzanian mining legislation has legitimatedforeign companies to acquire land previously worked infor-mally by small-scale miners, making small-scale mining-basedlivelihoods uncertain relative to foreign mining interests inTanzania (Fisher, 2007).

(a) Evolving organizational hierarchy of small-scale mining

In Tanzania, PML claim owners are responsible for miningactivities conducted on their claim, including hiring and pay-ing labor, organizing the mining, and endorsing safety andenvironmental regulations (Fisher, 2007; Mwaipopo et al.,2004). But PML owners’ active involvement in these mattersrarely transpires given the history of small-scale mining. WhenAfricans were initially granted mineral rights toward the endof colonial time, the geological uncertainties and expensesassociated with mining meant that a system of informalsub-contracting and shareholding emerged (Chachage, 1995).Today, the same principles of risk and profit sharing formthe basis of most small-scale mining activities. The actualsystem depends on local conditions of labor availability, reefrichness, and labor skills. However, in most mining sitesa three-tier division of labor has developed between PMLowners, pit holders, and diggers (Figure 1).

A PML grants the right to exploit an area for 5 years. Thearea can be mortgaged, renewed, or transferred to anotherholder, including foreign firms. It is a widespread practiceamong PML owners to informally lease out mining activitiesto pit holders, who organize procurement and sourcing of nec-essary inputs and labor and conduct the mining. Thus, capitalinvestments and the risks and costs related to fruitlessperiods fall upon pit holders, who often run their activitiesin partnership with others, hence their local name of wanachama

Localized Mining as Income Diversification

PIT HOLDER

risk-bearing investors/ site managers

CLAIM OWNER legal PML licensee

Mobile Mining Fraternity

DIGGERScontracted team workers

PROCESSORS task-performing casual workers

Figure 1. Gold production organizational structure.

382 WORLD DEVELOPMENT

(members). Typically, diggers are provided only with food andmedicine in periods of no mineral output. They can roughly bedivided into traditional drillers, waponjaji, who work their wayinto the rock mainly with hammers and chisels, and vutafelo,who remove the waste material and gold-bearing rocks outof the pit. The former usually receive a larger share than thelatter. In addition, small-scale mining sites have people occu-pied with various specialized tasks for which they are typicallypaid a specific amount of cash or quantity of rocks by task, forexample, security, pit supervision, blasting, electrical drilling,ore transportation, manual ore crushing, operation of ballmills, and separation of gold through washing and amalgam-ation.

Systems vary between mining settlements. Overall, claimowners typically take around 30% of the output, the pit hold-ers’ share adds up to around 40%, out of which productioncosts are usually covered, and diggers receive the remainingthird to divide among themselves (Jønsson & Fold, 2009).Sub-leasing of pits contradicts the mining legislation and if aPML claim owner chooses to sell his license, pit holders anddiggers are likely to be evicted. Thus, internal forms of exploi-tation exist within the small-scale mining sector with claimowners topping the pyramid (Fisher, 2007).

(b) Losing out to large-scale mining

Tanzania’s mining legislation has attracted some of theworld’s largest mining companies plus a large number of med-ium-scale prospecting and mining firms. Since 1997 foreign di-rect investments in mining have risen to over US$ 2.5 billionand gold exports have increased from constituting under 1%of export revenues in the late 1990s to 42% in 2005, whenthe country became Africa’s third largest gold exporter (Mac-donald & Roe, 2007). Nevertheless, observers question the so-cio-economic benefits of large-scale mining (Butler, 2004;Emel & Huber, 2008; Lange, 2006). While Tanzania’s gold ex-ports mainly from large-scale mining in 2005 were worth US$640 million, only US$ 30 million was collected in governmentrevenues (Emel & Huber, 2008). And in 2004, only 8,000 peo-

ple were employed within the large-scale mining sector, a num-ber that corresponds to only a very small fraction of the small-scale mining labor force (Macdonald & Roe, 2007).

Large and medium-scale prospecting and mining companieshave been buying up small-scale miners’ PMLs, in some casesinstigating the eviction of hundreds of small-scale miners.Generally, it has proven difficult for people to acquire PMLs,primarily because large tracts of mineral-rich land in the tradi-tional mining districts have already been acquired by commer-cial companies with Prospecting Licenses (PLs). 5

With the rapid expansion of commercial mining, access toTanzania’s mineral wealth seems increasingly to be in thehands of predominantly foreign-owned companies. Thus, con-temporary small-scale mining is experiencing a boom phase,which is likely to be a restricted short-lived window of oppor-tunity for work, careers, and investment. As the price of goldrises on world markets amidst global financial upheaval, it re-mains to be seen how small-scale miners will fare vis-a-vis thepressures of large-scale mining in Tanzanian gold. This is thebackground context of our case study focus on Tanzaniansmall-scale miners’ careers.

4. GOLD DIGGERS EMERGING FROM THELIVELIHOOD SCRAMBLE

While the literature on African small-scale mining has beenexpanding rapidly over the last decade, there is surprisingly lit-tle on small-scale miners’ decision-making and career trajecto-ries. With the aim of studying the work and mobility patternsof small-scale miners, we designed a field study focussed onminers of different ages and degrees of mining experience intwo contrasting gold sites.

(a) Methodology and study areas

Our survey took place in the gold mining settlements ofMatundasi and Londoni for 11 months intermittently duringAugust 2006–February 2008. 6 The village of Matundasi, lo-

GOLD DIGGING CAREERS IN RURAL EAST AFRICA: SMALL-SCALE MINERS’ LIVELIHOOD CHOICES 383

cated in Mbeya region in southwestern Tanzania, has been anactive mining settlement since the 1930s (Roberts, 1986). Thelegalization and growth of small-scale mining and recent risein the price of gold have, however, transformed the villageover the last two decades. In the early 1980s, Matundasi hadonly one small shop, making it necessary for residents to travel100 km to Mbeya town for most supplies; at the time of oursurvey it had 38 shops, 26 eating places, 12 bars, and six hairsalons. According to the latest village census of 2007, Matund-asi has a population of 7,640, of whom an estimated 1,500–2,000 are small-scale miners who mine as their sole livelihood.Of these, however, only nine own PML claims, as most land inMatundasi is held as PLs by commercial companies.

Londoni is a village in Singida region in central Tanzania,which experienced a sudden gold rush in July 2004. At thetime, the village had 1,600 residents involved in subsistencefarming and livestock keeping and two small shops. A fewmonths later, more than 10,000 people were estimated to beengaged in gold mining. During the study, Londoni had,among other things, 74 shops, 47 eating places, 29 bars, andnine hair salons. In early 2008 there were 215 PMLs in Lon-doni, as miners managed to secure a large area for small-scalemining prior to the 2005 national elections by putting pressureon key politicians. With a more settled, family-based popula-tion, Matundasi had proportionately fewer young men onhand at the time of the study compared with the rush settle-ment of Londoni, where there were many experienced, mobileminers between 20 and 39 years of age. In both areas, therewere few miners over 50, indicating the heavy physical de-mands of mining and the comparatively short-lived natureof mining careers relative to rural farming.

Many small-scale miners are wary and resistant to outsiders.Some are involved in illicit forms of mining. Others are suspi-cious of potential rivals in pursuit of mineral-rich land. Thus,creating trust and rapport with miners was vital. This wasdone by getting familiarized with the communities throughregular meetings and participation in a number of social activ-ities. 7

(b) Profiling the miners

Biographical and socio-economic profiles of small-scaleminers are foundational to understanding the miners’ motiva-tion and decision-making behavior (Hilson, 2007). All of our108 survey respondents were men. In Matundasi and Londoni,female miners did not appear in great numbers. Male minersargue that women entering the pits bring bad luck. The few fe-male mine workers encountered throughout the fieldworkeither were occupied with crushing, washing, panning, andtransporting of sediments, or (in a few cases) had shares inPMLs or pits.

The men in our survey represented 27 of Tanzania’s approxi-mately 125 tribes, originating from 15 of the country’s 21 main-land regions. Characteristically mobile, they had chalked up 293stays at 70 different mining sites, spending an average of 41months per site and working a mean of 2.7 sites. Of the 108respondents, 24 were either claim owners or pit holders. Atthe top end of the mining pyramid, they were more mobile thandiggers with 3.8 site stays each (Jønsson & Bryceson, 2009).

A digger’s first mining site is effectively an apprenticeshipwhere he learns how to mine and socially interact within themining settlement. Some are reluctant to move away from thissite, which is often located close to the home area, but manygradually build up the confidence and skills to move on. Ca-reer aspirations begin with the commencement of work at asecond site. From here onwards, miners spend on average less

time than at the first site, as they focus on the material rewardsof mining and leave when deposits are depleted or a newpromising site is discovered (Jønsson & Bryceson, 2009). Asminers move from site to site, they increasingly distance them-selves from their home areas, thereby weakening their familyties, but they compensate by forming attachments to house-holds in mining sites or other localities. Individual miners de-pend on each other in very systematic ways pertaining to roles,obligations, and shares. Miners’ careers are essentially moldedby how they manage to access gold-bearing land, gold, miningteams, equipment, mining knowledge, markets, security, andcredit. On arriving at a new site, miners need inter-personalskills to negotiate effective social networking, which is crucialfor gold mining access and success.

Claim owners tend to rely on relatives to manage their oper-ations when they are absent from the claim. Outside of thisproperty guardianship function, the importance of familialand tribal ties does not seem salient in mining. Ethnic and tri-bal boundaries are intersected by new cultural, political, andtenurial relations within the mining settlements (Godoy,1985). Miners are increasingly judged for their qualificationsand competence instead of place of origin and tribal affiliation.Consequently, although some pit teams seemed to be formedon a tribal basis, these constituted a small minority. Seventy-seven percent of our respondents originated from rural areas,which is exactly the same percentage reported for rural men inthe most recent national census (Tanzania, 2003), indicatingthat mining is as attractive to urban dwellers as it is to ruraldwellers.

Small-scale mining is an open-entry occupation involvingpeople from youth to late middle age. Our respondents startedmining during 1974–2006. Some youths began as early as 12years old, but on average, miners’ starting age was 24. To gaindetailed insight into miners’ work trajectories and comprehendthe significance of timing in embarking on a mining careerwith respect to entry year and starting age, we have con-structed three different category groupings. The ‘‘mining entryyear cohorts” (Table 1) reflect the responsiveness of miners tochanging government policies toward mining and gold strikeinformation. The ‘‘career entry age cohorts” (Table 2) are con-structed by comparing individual miners’ entry age relative totheir mining experience in years. Those who start at a youngerage than the average age are grouped as ‘‘early starters,” fol-lowed by ‘‘average starters,” ‘‘delayed starters”, and finally‘‘late starters.” The aim is to shed light on miners’ career deci-sion-making with respect to the progression of their life cycle.We hypothesized that those who start early and persist aremore likely to succeed than latecomers. Finally, we groupedthe miners with respect to their hierarchical function (Table3)—diggers, pit holders, and claim owners—with the expecta-tion that claim owners at the top of the pyramid will have thehighest material rewards.

Table 1 shows that from a very low number of entrants be-fore 1985, there was an almost fivefold increase in the numberof men starting mining in the late 1980s, coinciding with thegovernment’s initiation of economic liberalization policies.Thereafter a steady stream of men entered mining, especiallyin their early to mid 20s. Although miners with many yearsof mining experience have worked the most sites, there ap-pears to be a general trend for the duration of site stays to de-cline with later mining entry cohorts, hinting at increasedinter-site mobility. The large cohort in 1985–89, which regis-ters comparatively low mobility, is less educated than anyother cohort. Those were years of economic distress and it ispossible that many of these men entered mining as a copingrather than an optimizing strategy. The respondents who

Table 1. Career characteristics of ‘‘mining entry year” cohorts. Source: Survey data

Cohorts No. ofrespondents

Averageage

Averageentry age

No. of sitesworked

Years of miningexperience

Mean monthsper site

Less than 5years of school (%)

Ruralorigin (%)

Originating fromtraditional mining

areas (%)

Pre 1979 5 49.6 21.0 5.0 18.4 44 20.0 80.0 60.01980–84 5 51.8 27.8 5.2 17.0 39 20.0 100.0 100.01985–89 23 46.3 26.4 3.2 16.0 60 34.8 78.3 56.51990–94 19 38.1 23.7 3.5 11.3 39 5.3 47.4 73.71995–99 18 32.4 22.9 2.7 7.9 35 22.3 83.3 44.42000–04 19 26.5 21.6 1.6 4.5 33 26.4 73.7 36.82005+ 19 26.7 25.4 1.2 1.1 11 26.4 94.7 78.9

Total 108 36 24.2 2.7 9.3 41 23.2 76.9 60.1

384 WORLD DEVELOPMENT

started mining in 1990–94, on the other hand, stand out asmore urban and better educated than any other cohort, withthree-quarters coming from traditional mining areas.

Table 2 reveals two, rather than one, career paths in mining.The ‘‘early starters,” ‘‘average starters”, and ‘‘late starters”display a normal progression path, that is, the earlier theystarted mining, the more sites they worked and the longer theystayed at each site. There is a logical pattern of a decreasingnumber of sites and decreasing duration at each site for theaverage and late starters. The late starters spent approximately1 year less at each site compared with early and average start-ers. And in general, the later the entry, the less time spent ateach mining site before moving on. This could be a result ofboth increased pressure from commercial mining companiesand the fact that the legalization of small-scale mining gaveminers more freedom of movement to search for the most pro-ductive mining sites.

The delayed starters, however, deviate from this pattern.They were 5–10 years older than the early and average starterswhen they began mining but they ‘‘geared up” very quicklyand managed to work at more sites than average startersand spent far shorter periods at each site than any of the othercohorts. Table 2 shows that the overwhelming majority ofthem are from traditional mining areas, which may help to ex-plain how they adjusted so quickly to mining. Furthermore,they tend to be the least rural of any of the groups. We willreturn to the significance of these characteristics later.

Distinguishing miners by functional hierarchy in Table 3, wesee that all three groups have almost identical entry ages, buttheir current ages differ. Pit holders and claim owners are 8–9years older than diggers, inferring that work experience is req-uisite to climbing the pyramid to its second and third tiers. Aprogression up the functional small-scale mining hierarchy ex-ists. All but two of the 24 pit holders and claim owners startedtheir mining careers as diggers and managed to work their wayup to their current position.

Pit holders have a higher mobility than any other group.They have visited more sites and spent less time on each site.However, there is a big difference in mobility between miners:miners from Matundasi have an average of 2.3 site visits whileminers from Londoni have an average of 3.1 site visits. Thefour claim owners, on the other hand, were the least mobile,averaging 7.8 years per site, which accords with the fact thattheir ownership rights require a more sedentary existence closeto their site for supervision purposes. If a new gold rush oc-curs, claim owners are more confined to their claim, for whichthey hold the legal title, than are pit holders to their existingpits. Education level may facilitate advancement: claim ownersare better educated than the pit holders, who are in turn bettereducated than diggers. Pit holders, and especially claim own-ers, are more likely to have urban origins and come from tra-ditional mining areas compared with the diggers.

5. PROFESSIONAL CAREER TRAJECTORIES WITHINSMALL-SCALE MINING

Compared with the multi-various ad hoc non-farm rural activ-ities that have proliferated over the last few decades, small-scalemining has a readily identifiable career structure as outlined byKanter (1989). The sector presents miners with the opportunityfor career progression from ‘‘early apprenticeship” to successivemining sites, requiring readiness to move, dedication to thework, and good interaction with fellow miners.

This marks a step away from the non-farm employmentexperimentation of the last three decades in Tanzania, whichwas propelled by people trying to cope with material adversityand ensure physical survival (Bryceson, 1999). It was in thiscontext that people began experimenting more with mining.Mining was seen as a temporary employment that people pur-sued only until they earned sufficient money to go back tofarming (Madulu, 1998). However, a growing disinterest withfarming as a way of life, particularly among the younger gen-eration, has surfaced and today, increasingly more peopleidentify themselves occupationally as miners. The SAP re-forms of the 1980s pushed farmers into the non-farm sectorfollowed by economic liberalization in the 1990s and the legal-ization of small-scale mining, providing people with theopportunity for a permanent mining-based livelihood insteadof the ‘‘hide (from the authorities) and seek (minerals) life”miners had been living until then.

Less than 5% of our respondents mentioned post-miningagrarian strategies. Business initiatives, commercial mining,and consumption figured prominently as miners’ ultimate sav-ing goals (Jønsson & Bryceson, 2009). It is increasingly appar-ent that multi-sited small-scale miners are no longer frustratedfarmers turned miners for survival reasons. On the contrary,we are witnessing the emergence of career miners with min-eral-led spatial mobility strategies, social mobility aspirations,and collective identity characteristics.

Miners accept continued material hardship and separationfrom family as part of their work. While their initial entry intothe sector does not entail significant barriers, sustaining a ca-reer within small-scale mining involves enduring a number ofhazards and uncertainties of livelihood failure, family separa-tion, and occupational risks. Conditions in mining communi-ties can be harsh and miners have to be resistant to adversityand persistent in the pursuit of their career. Many quit afterexperiencing economic ruin, social demoralization or poorhealth or witnessing mine accidents. Despite the high intakeof alcohol and smoking of marijuana, especially among youth-ful diggers, most small-scale miners are committed to hardwork and display a strong work ethic.

Barret, Reardon, and Webb (2001, p. 325) argue that ‘‘sub-stantial entry or mobility barriers to high return niches withinthe rural non-farm economy” exist. However, although small-

Table 2. Characteristics of ‘‘career entry age’’ cohorts.a Source: Survey data

Cohorts No. ofrespondents

No. of sitesworked

Years of miningexperience

Averageentry age

Currentaverage age

Mean monthsper site

Less than 5 yearsof school (%)

Ruralorigin (%)

Originating fromtraditional mining areas (%)

Early starters 28 3.4 13.7 17.4 34.0 48.1 21.4 82.1 46.4Average starters 47 2.5 9.2 22.2 33.3 43.4 29.8 74.5 66.0Delayed starters 19 2.9 6.8 27.2 37.8 28.0 15.8 73.7 84.2Late starters 14 1.6 4.6 40.2 46.5 33.6 14.2 78.6 35.7

Total 108 2.7 9.3 24.2 36 41.3 23.2 76.9 60.1a In analyzing the career progression of men of widely varying ages in our sample, we grouped miners by their years of mining experience relative to their mining entry age. Those in the modal group were‘‘average starters,” compared with ‘‘early starters,” later ‘‘delayed starters”, and very ‘‘late starters” as follows.

Mining Experience (number of miners)

Current Age <5 years 5-9 years 10-14 years 15-19 years 20-24 years 25+ years

<20 14 4

20-29 5 7 6 3

30-39 4 6 10 3 2

40-49 2 2 4 9 6

50 + 1 1 4 4 7 4

Early starters = 28 above averageAverage starters = 47 modal averageDelayed starters = 19 below average

Late starters = 14 very below average

GO

LD

DIG

GIN

GC

AR

EE

RS

INR

UR

AL

EA

ST

AF

RIC

A:

SM

AL

L-S

CA

LE

MIN

ER

S’

LIV

EL

IHO

OD

CH

OIC

ES

385

Table 3. Career characteristics of ‘‘functional mining’’ cohorts. Source: Survey data

Cohorts No. ofrespondents

Average age Averageentry age

No. of sitesworked

Years ofmining experience

Mean monthsper site

Less than5 years

of school (%)

Ruralorigin (%)

Originatingfrom traditionalmining areas (%)

Claim owners 4 43 26 2.0 15.4 93 0 50 75Pit holders 20 42 25 4.1 11.4 33 10 65 75Diggers 80 34 24 2.4 8.6 43 28 80 55

386 WORLD DEVELOPMENT

scale mining is potentially one of the most lucrative non-farmactivities, its initial barriers to diggers’ entry are relatively lim-ited in terms of required investments and skills. Estimates fromPhillips et al. (2001) found average incomes within Tanzaniansmall-scale mining to be six times higher than in agricultureand Lange (2006) estimates average incomes for small-scaleminers in Tanzania to be approximately 120 US$ per month.Our rough estimate of respondents’ monthly income averaged158,000 Tanzanian shillings (Tsh), 8 supporting the argumentthat income levels within small-scale mining as a whole are high-er than those in farming. Respondents, covering a wide incomespectrum, earned from close to nothing up to one and a half mil-lion shillings monthly. Large differences between various co-horts exist with regard to performance and satisfaction(Tables 4–6). Monthly earnings increase with age up to 40 yearsand thereafter decline, which is likely to be indicative of decreas-ing amounts of time spent actively digging as the men age.

Excluding the ‘‘delayed starters,” we see a logical career pro-gression and the importance of early mining entry. The ‘‘early

Table 4. Performance of ‘‘career entry

Cohorts Mean monthlyearnings

(Tsh ‘000)a

Mean miningcareer earnings(Tsh million)b

Mining sitesummaryoutcomec

Mining’on livel

Early starters 152.3 25.0 2.91 1.9Average starters 125.2 13.8 2.78 1.6Delayed starters 276.9 22.6 3.15 1.8Late starters 114.5 6.3 2.77 1.5

Total 157.5 17.6 2.88 1.7a Given miners’ reticence about earnings, this figure is estimated indirectly thrfrom your earnings? Its Tsh cost? How long would it have taken to pay for frb Mean monthly earnings multiplied by total number of months worked per inc Based on the question: ‘‘How would you rate what you got out of the site?”good” 5.d Based on the question: ‘‘Overall, how has mining affected your livelihood?”

Table 5. Performance of ‘‘mining entry

Cohorts Mean monthlyearnings

(Tsh ‘000)a

Mean mining careerearnings

(Tsh million)b

Mining sitesummary outcomec

Mininliv

Pre 1979 211.8 43.7 2.821980–84 293,1 41.0 2.921985–89 149.6 29.0 3.021990–94 239.6 32.1 2.961995–99 155.4 15.9 2.742000–04 144.5 8.3 2.952005+ 50.0 0.7 2.68

Total 157.5 20.0 2.88a See Note a, Table 4.b See Note b, Table 4.c See Note c, Table 4.d See Note d, Table 4.

starters” are materially better off and more satisfied with theirsituation than the ‘‘average starters,” who in turn have higherearnings and are more satisfied than the ‘‘late starters.” Nota-bly, however, the ‘‘delayed starters” received substantiallyhigher monthly earnings than all other cohorts. They are rela-tively well educated, several are pit holders, and 16 of the co-hort’s 19 respondents worked in Londoni, where minersearned significantly more than the miners from Matundasidue to the recent discovery and richness of the Londoni golddeposits. Moreover, as shown in Table 2, it is the cohort withthe highest degree of mobility.

Familiarity with small-scale mining, obtained by peoplegrowing up in mining areas is an advantage. Eighty-four per-cent of the ‘‘delayed starters” come from traditional miningareas as opposed to only 36% of the ‘‘late starters.” Some ofthe ‘‘late starters” may have been pushed by livelihood failureinto traveling a long distance to start mining. Our data indi-cate a pattern of improving material reward and career pro-gression as miners become more experienced, skilled, and

age’’ cohorts. Source: Survey data

s effectihoodd

Claimowners (%)

Pitholders (%)

Diggers(%)

Matundasi/Londonirespondent distribution

3 2 32 61 18/106 2 9 89 27/209 32 68 3/160 7 7 86 6/8

5 4 19 78 54/54

ough questions regarding purchases: What big thing did you want to buyom your monthly earnings?dividual.

with the answers: ‘‘very poor” 1, ‘‘poor” 2, ‘‘OK” 3, ‘‘good” 4, and ‘‘very

with the answers: ‘‘very negatively” 1, ‘‘negatively” 2, ‘‘positively” 3, and

year’’ cohorts. Source: Survey data

g’s effect onelihoodd

Claimowners (%)

Pitholders (%)

Diggers (%) Matundasi/Londoni respondent

distribution

2.20 20 40 40 3/22.00 40 60 2/31.70 22 78 16/71.89 11 42 47 6/131.67 6 11 83 12/61.63 5 95 11/81.68 100 4/15

1.75 4 19 78 54/54

Table 6. Performance of ‘‘functional mining’’ cohorts. Source: Survey data

Cohorts Mean monthlyearnings (Tsh ‘000)a

Total miningcareer earnings(Tsh millions)b

Mining sitesummary outcomec

Mining’s effecton livelihoodd

Matundasi/Londonirespondents’ distribution

Claim owners 230.0 42.5 3.00 2.25 3/1Pit holders 311.0 42.5 3.10 2.05 4/16Diggers 117.0 12.1 2.79 1.65 33/47

a See Note a, Table 4.b See Note b, Table 4.c See Note c, Table 4.d See Note d, Table 4.

GOLD DIGGING CAREERS IN RURAL EAST AFRICA: SMALL-SCALE MINERS’ LIVELIHOOD CHOICES 387

mobile, and expand their networks within the mining commu-nity. However, alongside this pattern of career progression,the geographical origin and educational level of miners areimportant factors of success. The ‘‘delayed starters” have anadvantageous sideways entry into mining, which pegs themhigher than most others, facilitated by their locational advan-tage. They have grown up in mining regions and more than aquarter originate from urban areas, which may afford them abetter income-gathering position regarding gold strikes, facil-itating arrival at gold strike sites in a timely fashion that canmake a large difference to earnings (Jønsson & Bryceson,2009).

Miners with long mining careers out-perform those whohave recently entered mining, demonstrating the value of amining apprenticeship and accumulated work experience.Miners who started mining in the new millennium generallyearn less than other miners. But it is not just years of experi-ence that matter. Members of the large mining entry cohortof 1985–89 who responded to the exceptionally poor economicconditions of the time and may have been ‘‘driven to mining”are relatively poorly educated. Only 57% of them originatedfrom traditional mining areas. The majority of them work inMatundasi, where proceeds are generally lower than in Lon-doni, and tellingly, all arrived late at their last gold strike site.They were less satisfied with their situation compared to othercohorts. By contrast, miners in the 1990–94 mining entry co-hort, which was the highest earning cohort, were overall themost satisfied with their situation. Miners in this cohort werethe best educated. A high percentage of them originated fromtraditional mining areas and were mining in Londoni.

Finally, turning to the three functional groups, gold findingsare unequally divided between the claim owners, pit holders,and diggers. As expected, claim owners and pit holders earnmore and report more successful mining and livelihood out-comes than diggers. Interestingly, pit holders report earningsat parity with claim owners and slightly better mine site out-comes than them. This may be because claim owners are dif-ferentiated on the basis of the perceived gold-bearing qualityof their claim, whereas pit holders have inter-claim mobility.A pit holder on a PML with rich deposits can easily earn moremoney than a claim owner whose area only has a few low-yielding pits. Many claim owners with mediocre claims strug-gle to get pit holders to work on their claims because the pitholders prefer the PMLs located on rich gold reefs. 9

The two cases below illustrate the upward career progres-sion through the functional groupings as well as the impor-tance of timely movement, apprenticeship, and technologicalinvestments and innovation.

Fifty-year-old Thomas S. started mining in Itilima inShinyanga in 1982 at the age of 25. He worked as a diggerin other people’s pits and sometimes during the rainy sea-son also did alluvial mining. After 4 years he went to Mat-inje in Tabora, again being employed in other people’s pits.After 5 years, he went to Mwagi Magi in Shinyanga, where

gold had recently been discovered. After working in otherpeople’s pits for nearly a year, he went on to Sekenke inSingida. By then, he had learned his trade and upon arrival,he instantly started his own pit, which he operated as a pitholder for the next 7 years. When gold was discovered inLondoni in 2004 he obtained a pit. Nine months later thegold fever was on at Haydom in Manyara and he wasone of the first to arrive. Again he had his own pit. Butthe produce was limited and after a year, he was back inLondoni operating a new pit (Interview January 16, 2007at Londoni).

Forty-year-old Peter B. started mining as a 12-year-oldyoungster at home in Ushirombo in Shinyanga. Whileschooling, he mined for gold in his spare time, either allu-vial mining with friends or in other people’s pits. After fin-ishing school he became a full-time hard rock digger. Whengold was struck in neighboring Mwabomba, he went andstarted his own pit. After eight months, the gold-bearingreef became increasingly difficult to access. He bought aball mill from his mine earnings and transported it the700 km to Matundasi. Here, the ball mill was the first ofits kind, hence earning him good money from crushingother miners’ rocks. After a while he started operating hisown pit (Interview January 11, 2007 at Matundasi).

The fact that Peter B. came from a traditional mining areaand had 11 years of schooling is likely to have played a signif-icant part in his skills acquisition and acumen to invest in pro-cessing equipment fairly early in his mining career and maketimely moves to new mining sites.

6. FRATERNITY OF ‘‘REAL MINERS’’: CAREERSTRUCTURE AND PROFESSIONAL IDENTITY

Stress so far has been primarily on career entry, now we turnto the internal dynamics within the mining hierarchy and thedelineation of career miners from ‘‘others” working in goldmining. Gold excavation is organized into production shifts. 10

Pit holders, sub-contracted by the claim owner, are central toputting together the shifts. They recruit the labor and mobilizewith the purpose of excavating gold from the mining shafts(Jønsson & Fold, 2009). The shift work is considered to bethe potentially high-yielding core of small-scale hard rock min-ing, which claim owners, pit holders, and diggers focus theirattention on in gold rush settings.

Less remunerative processing work is generally sub-con-tracted, for example, transportation, crushing, and grindingof rocks, washing, and amalgamation involve large numbersof casual workers who are hired on a daily basis. These are of-ten referred to as wasalalaji. Their numbers are augmented byalluvial panners working on their own account. It is usuallyfrom the ranks of casual gold processors and panners that dig-gers are recruited into production shifts. Many processers,mainly women, the elderly, and men lacking physical strength,

388 WORLD DEVELOPMENT

may never have the opportunity to move into what might bestbe called the ‘‘mining fraternity.” On the other hand, strongyoung men may bypass gold processing tasks and be directlyrecruited to the arduous task of hoisting rock out of miningpits.

There is a tacit divide between people engaging in gold pro-cessing and the pit-based work teams. Many of our experi-enced respondents referred to themselves as wachimbajihaswa (genuine or ‘‘real miners”) or wataalam (mining ex-perts). It was their way of setting themselves apart as special-ized workers who shared an agreed division of labor andproduction shift organization. ‘‘Real miners” not only workedtogether, but also played together. They frequented specificbars and recreational centers after work where they socializedand exchanged work-related information. Above all, ‘‘realminers” prioritized the search for and production of gold intheir lives, willing to follow the gold wherever that might leadthem. Gold processors, on the other hand, tended to be con-sidered on-hand local assistants doing processing work toboost their cash income. These are village men, women, andyouth who generally are otherwise engaged in trade or farm-ing. Figure 1 shows a divide between the fraternity of ‘‘realminers” and gold processors.

It should be noted, however, that the ‘‘real miners” may‘‘sink” into the processing pool between shifts, when theyare down on their luck and short of cash or during the rainyseason where tailings are reworked. Thus, there are backwardand forward linkages between the two worker categories aspersonnel slip from one to the other. Nonetheless, everyoneis cognizant of the corporate integrity of the ‘‘real miners,”who are propelled by career commitment to mining overattachment to place of residence.

Upward mobility within the mining fraternity takes theform of enterprising diggers who have managed to savemoney for capital investment and may become pit holdersor claim owners. Rising from digger to pit holder involvesa willingness to take on higher levels of risk, investing moneythat may be lost if gold is not found. Moving from pit holderto claim owner is not necessarily a straight-line option.Obtaining a gold claim generally involves being in the rightplace at the right time. Miners’ social networks are rife withrumors about claim opportunities. Having the right politicalconnections is a bonus. Inheritance is another point of claimentry. Miners die from accidents and diseases includingAIDS, leading to their next of kin inheriting their claim. Wo-men claim owners tend to come into gold mining throughthis entry point.

Given the uncertain character of gold finds, no matter howskilled and how carefully one manages one’s investments, longstreaks of not finding gold can push people down in the hier-archy and sometimes out of mining into farming. When ca-reer-minded miners sink to these levels or are forced to takeup farming for survival they are not positioned at the centerof ‘‘real mining” and risk-taking for lucrative earnings. Thisis reflected in the following quotes from pit holders and claimowners.

The most casual labor force employed in washing or basichand crushing activities or as normal mine workers (vuta-felo) may go home to farm. Real miners do not leave togo farming. (Hamisi M., pit holder, Londoni, February11, 2008).If the real miners want to farm, they send money home totheir families so they can hire labor. (Gasper T., pit holder,Londoni, February 10, 2008).Miners don’t waste time on farming. (Masaka S., pitholder, Londoni, February 11, 2008).

The real miners rarely farm. . .[but a] farmer who is just inthe mining settlement to get some cash for consumptionand to run his farming activities, he may go away to farm.This will sometimes create a minor shortage of [mining]labor.” (Boniface N., claim owner, Matundasi, February4, 2008).Miners don’t farm unless they have not been gettingenough in the mines. Once they get gold, they don’t thinkabout farming. (Thomas I., pit holder and gold broker,Londoni, February 11, 2008).

Attitudinally, the ‘‘real miners” feel farming is a sidetrackand to be a successful career miner it is best to disassociateoneself from farming to pursue mining seriously; this is almosta sine qua non of aspiring mining careerists. Thus, career-minded ‘‘real miners” distinguish themselves from local farm-ers as well as gold processors working on casual terms whogenerally have greater recourse to farming.

7. LOCATIONAL INFLUENCES ON MINERS’ CAREERCHOICES

‘‘Real miners” prioritize gold production and organize theirwork life accordingly. This requires being where the gold is.Given the fairly rapid mineral resource depletion of gold atlevels that small-scale miners can exploit, the mining popula-tion composition of mine site locations will vary over time,with some degree of out-migration as gold returns diminish.Our two survey sites represent different stages of the processof diminishing returns. Matundasi peaked in the last half ofthe 1980s in terms of incoming miners whereas Londoni hada gold rush and sudden influx of hopeful migrants in 2004. Gi-ven Londoni’s recent and targeted immigration of miners, weanticipated that Londoni’s miners would evidence stronger ca-reer specialization in gold production.

Table 7 shows that the surveyed miners have similar demo-graphic characteristics with respect to age, marital status, num-ber of children, and percentages originating from rural areas,but, as expected, miners in Matundasi have a more sedentaryprofile. More of them are indigenous to Matundasi or considerit home relative to the surveyed miners of Londoni. More ofLondoni’s miners come from Tanzania’s traditional miningareas. All indicators point to Londoni miners being more mo-bile with a higher number of work sites even though they areon average younger and have only spent 60% of the timeMatundasi miners have worked in mining. Finally, there is farless recourse to farming in Londoni relative to Matundasi. Lon-doni miners are more reliant on purchased food supplies with ahigher percentage of their expenditure devoted to food pur-chase. In Londoni, 18% of the miners (not including retirees)were engaged in non-mining business and trade activities andnon-farming activities, whereas in Matundasi, 33% of activeminers pursued such activities in combination with mining. 11

Our two survey areas reflect the very different opportunitiesto progress up the mining hierarchy. Matundasi is a maturesite where claim ownership for small-scale miners is highly re-stricted and monopolistic. There are only nine claim owners inthe area given the limited availability of PMLs. This contrastswith Londoni, where the national elections of 2005 catalyzedclaim acquisitions. Londoni’s gold strike in 2004 led tosmall-scale miners clamoring for claims. The ruling party’scandidates recognized the vote-winning potential of the situa-tion and managed to get the government to set aside 10 km2

for small-scale mining. By 2007, there were 215 PMLs in thearea (Jønsson & Fold, 2009).

Table 7. Comparison of survey miners’ characteristics by location. Source: Survey data

Matundasi (old mine site) Londoni (strike site)

Population of mine site 7,640 Approx. 10,000Miners from rural origins 78% 76%Married miners 57% 56%No. of children 2.6 3.1Miners’ average age 36 36Miners indigenous to site 31% 24%Miners considering ‘‘home” to be mine site 24% 2%Miners with traditional mining area origins 41 80Average no. of work sites 2.4 3.1Average no. of years mine experience 11.6 7.1Average no. of years per site 4.8 2.3Miners without farming activities 59% 91%Farming as significant non-mining activity 22% 6%Producing own staple food 15% 4%Engaging in non-mining/non-farming activities 33 18Mean annual income (total mine income/total years in mining) Tsh 1.8 million Tsh 2.7 millionMine site summary outcome 2.82 2.94Mining’s effect on livelihood 1.72 1.78

GOLD DIGGING CAREERS IN RURAL EAST AFRICA: SMALL-SCALE MINERS’ LIVELIHOOD CHOICES 389

Undoubtedly, not everyone had equal access to this goldenopportunity. Among Londoni claim owners and pit holders,the largest number were early starters (41%) followed by de-layed starters (35%), most of whom were migrants fromSukumaland, Tanzania’s main gold mining area. In Matund-asi, delayed starters were not in evidence. Early starters(57%) dominated, while late starters, indigenous to the region,accounted for 29%. Thus, our survey suggests that small-scalemining career progression is upwards through the ranks ofdiggers, but finding strike sites in a timely fashion helps someleap ahead in earnings and career prospects.

Finally, in concluding our consideration of the influence oflocation on mining careers, it is useful to indicate how small-scale miners’ pursuit of gold may not only shape a gold settle-ment’s population size and occupational structure over timebut also influence the ‘‘sedentary” as opposed to ‘‘mobile” nat-ure of the mining population and the relative portion of spe-cialized as opposed to diversified mine workers at the site.Figure 2 stylizes the interaction of mining, agriculture, andtrade in settlements experiencing a mining strike. In Stage 1,gold is discovered and is mined by a few local residents. Thelargely agrarian nature of the settlement is transformed inStage 2 as large numbers of migrants are attracted to the goldstrike site to mine in the first instance, followed by many menand women migrating to provide trade and services to thecomparatively cash-rich mining population. Sex ratios of thesettlement start to equalize, given the strong local economymultiplier effects of escalating gold production. In Stage 3,

Figure 2. Stylized sequencing of mining strike: population growth and

sectoral activity. Source: Based on Bryceson and Mwaipopo (2009).

small-scale miners are reaching depths at which it is difficultto extract the gold, prompting many miners to leave in searchof a new gold site. Their movement and career progression arethe focus of this article, whereas others, due to livelihood incli-nation or family ties, may choose to remain at the depletinggold site. They are likely to continue mining, supplementedby trade or agricultural activities. This phenomenon is rein-forced by the retirement of miners who seek to move into tradeand services as discussed in the next section. Thus, career-minded miners gravitate to Stage 2 settlements exemplifiedby Londoni, as opposed to a settlement exhibiting Stage 3or 4 characteristics like Matundasi. Table 7 illustrates this pat-tern. The more sedentary miners are found in Matundasi asopposed to Londoni with its large number of migrant ca-reer-minded miners.

8. CAREER MINERS’ RETIREMENT: SPENT FORCEOR SECOND WIND?

It is revealing to consider what miners do after they retirefrom mining. Bearing in mind that the vast majority areuntraceable, we, however, managed to interview 18 sedentaryex-miners in the two survey mining sites. On average, theywere 39 years old, had mined for 12.5 years at 2.9 sites, andwere generally positive about their mining careers. They wererelatively well educated with all but one having finished pri-mary school. All had capitalized on their earnings by investingin various kinds of local service provisioning or trading busi-nesses.

Having argued that career-minded miners avoid being asso-ciated with agriculture, it is not surprising to note that onretirement they aim to move beyond mining into businessrather than agriculture. Recognizing that the actual miningwork is a ‘‘strong man’s domain,” and that age will take itstoll, many successful miners see a switch to trade as their finalstrategic career decision. Ideally, they have managed to savesome of their gold earnings to invest in lucrative trade and ser-vice provisioning.

Their reasons for drawing their mining careers to a closewere wide-ranging and mostly connected with dissatisfactionwith mine work: unhealthy, hard, and risky work (45%), lossof money on mining (17%), lured to a new activity that paidbetter (22%), or simply tired of mining (17%). Many minersstop mining due to weariness. The majority (61%) stated that

390 WORLD DEVELOPMENT

life had improved economically since their retirement linked tothe increased stability and higher income of their currentwork, and the fact that some had been unlucky in mining be-fore making the decision to quit. The rest were phlegmaticabout their change in occupation, stating that there was littlechange in income level, but they appreciated that their presentearnings were more stable. Their slightly higher educational le-vel is likely to have given them the self-confidence to leavemining and strategically start a new occupation. Significantly,they had remained in the mining settlement where their knowl-edge of mining and social networks of the small-scale miningcommunity facilitated making a living from non-mining activ-ities. Most viewed their current sedentary life with satisfaction,although one third still regretted not being able to follow newmineral rushes.

Kiduta M., 38 years of age, stopped mining back in 2002when a pit in which he was mining collapsed: ‘‘I was half bur-ied in gravel and stones, but luckily people were close by. . .They got me up, but my knee got severely injured.” He hadbeen saving up and when he was able to walk again, heopened a clothes and shoe shop in Londoni. ‘‘It is too uncer-tain to dig for gold. . . One risks one’s life and limb. . . Today Iearn more money selling clothes and shoes than I earned inthe mines.” (Interview at Londoni, February 7, 2007).

Reflecting on their past mining careers, the ex-miners weredivided about whether mining was a worthwhile occupationto pursue. Only a third mentioned that it had afforded theman improved livelihood and 11% said that they had enjoyedthe excitement of traveling. But many had regrets. Over 40%felt that they had wasted their time and 6% stated that mininghad ruined their health. Another 6% felt that they had fritteredaway their earnings on mindless conspicuous consumptionand too much drinking. One third regretted that they hadnot left mining to do business earlier. However, half of therespondents were unequivocal about the fact that they wouldchose mining if they could start their work careers all overagain.

9. CONCLUSION

Gold mining in Tanzania is not a new phenomenon. A fewareas of the country have been sites of small-scale miningactivities for several decades, but gold mining has spread tovarious new parts of the country and gold has now gainedprominence as a major national export. This has been facili-tated by the rise in international gold prices and Tanzaniangovernment policies. As prospectors and producers, small-scale miners have played a central role in this economic surge.

Western literature on career structure tends to assume for-mal employment conditions such as those found in a civil ser-vice or corporation. But more broadly, careers encompass alifetime work pattern which can ideally be traced ‘‘onwardsand upwards” (Ip, 2008). It is in these terms that the conceptof career is relevant to many Tanzanian gold miners. Remark-ably, as small-scale mining has gained momentum within thelast two decades, small-scale mining career trajectories haveemerged despite miners’ uncertainty of finding gold in any gi-ven locality combined with their shifting mine sites, informallyconstructed work teams, and ad hoc subsistence payments. Ofthose who persist ‘‘onwards,” many rise ‘‘upwards.” The lat-ter’s career commitment is reflected in their willingness to con-tinually uproot themselves and dash to new gold strike sites.Experienced miners can be rewarded with progression fromdigger to higher earning pit holder and/or claim owner. In thisway, miners’ mobility, risk-taking, and tenacity to carry on

working for many years despite physically demanding laborand hardship pays off. Uncountable numbers leave, mostprobably after the initial mine site, whereas others drop bythe wayside finding alternative livelihood activities or areforced to stop mining due to injury or death. Mining accidentsand high HIV prevalence rates take their toll. But those thatstay are propelled by the possibility of being rewarded withlarger incomes as pit holders and claim owners. Climbing upthe career pyramid is facilitated by one’s origins near to a min-ing area, a comparatively higher education level, a willingnessto move from site to site, and arriving early at strike sites(Jønsson & Bryceson, 2009).

In this article we have argued that Tanzania’s small-scalemining sector represents a site of coalescing career formationarising almost entirely from the small-scale miners’ own orga-nizational constructs and individual decision-making. Wehave identified a path of strategic choices propelling mineworkers toward labor specialization. In the first instance, theyseek mine work as income diversifiers extending beyond tradi-tional agriculture or more recent burgeoning trade activities.This may require moving to a mine site if they are not alreadyresident there. In most cases they join the ranks of gold proces-sors and from there some of the men are recruited into the‘‘mining fraternity” working as part of a digging team on averbal contractual basis under the direction of the pit holder.Working in this capacity for a couple of years, building upskills, and social networks among other members of the min-ing fraternity, and/or proceeding to work at another miningsite in the pits, constitutes the beginning of an individual dig-ger’s mining career. The digger embarks on a tangible careertrajectory and material incentives that propel him to othersites, where his collegial ties and skills can further expand.

In small-scale mining studies, there is a need to distinguishbetween income diversifying mineworkers aimed at povertyalleviation and small-scale miners following a career trajec-tory, which affords them the means for material advancementof their lives. However, because of the extremely high risksand uncertainty of small-scale gold mining in rural Africa, ca-reer miners are rarely divorced from the possibility of havingto relinquish their career specialization to become generalrural income diversifiers experimenting with combiningactivities—farming, trading, service provisioning, andmining—into viable livelihoods.

Tanzania’s small-scale miners are often criticized for work-ing on illegal mining sites and in illicit gold trade, engagingin heavy drinking and prostitution, and causing environmentaldegradation. They are frequently viewed by local farmers andgovernment officials as misfits. On the other hand, their activ-ities have a dynamic impact on the local economy and theirwork is highly labor absorbing and productive. Their careerdedication enhances their professional competence in an areaof work with potentially high material rewards, affording themthe means to anticipate, plan, and invest over the course oftheir work lives (Kanter, 1989). This contrasts with the contin-uing amorphous nature of self-employed income diversifica-tion more generally with its uncertainty of markets, supply,and infrastructure.

Mining opportunities at the base of the career pyramid af-ford the entry of massive numbers who try their hand at min-ing. Those who gain the skills and have the determination tomine believe with good reason that they will earn a livelihoodthat is likely to improve over time. They form the role modelsand upper cadre who maintain the coherence of Tanzania’ssmall-scale miners’ career hierarchy.

But the question is how long this will last. Small-scale min-ers face technical limits to the depths to which they can dig.

GOLD DIGGING CAREERS IN RURAL EAST AFRICA: SMALL-SCALE MINERS’ LIVELIHOOD CHOICES 391

Where mineral wealth is abundant, large corporate interestsinevitably edge out small-scale producers. Tanzania is noexception to this pattern. The government has, in conjunctionwith the World Bank, put legislation in place to encouragelarge-scale mining investment. This calls for thinking aheadand devising policies for small-scale miners who become mar-ginalized (Hilson et al., 2007).

The Tanzanian gold diggers, pit holders, and claim ownersthat we have focused on in this study have demonstrated per-sonal initiative and collective creativity in evolving a careertrajectory and professionalism, enabling them to be productiveto the benefit of their families and the country as a whole.Their pyramidal career hierarchy has succeeded in providing

them with valuable work skills and considerable self-gover-nance. As large-scale mining increasingly encroaches onsmall-scale mining areas, the onwards and upwards careermovements of small-scale miners will be stymied and demoral-ization will set in. Policy measures should be devised andimplemented in a timely fashion to preclude penalizing anexceptionally hard-working and creative segment of the Tan-zanian workforce. Their prospecting work has pointed theway for large-scale mining. As Tanzania’s mineral wealth ac-crues, the government and large-scale mining firms would beadvised to design schemes to recruit labor and encourage ser-vice sector employment from the ranks of these path-breakingminers.

NOTES

1. Defined here as individual or collective labor-intensive mineralextraction, legal or illegal, with limited capital investments using basictools, manual devices, and/or simple portable equipment.

2. This refers to miners in African countries experiencing a succession ofmineral discoveries.

3. There are, however, a number of studies which provide importantinsights, especially in contexts where livelihood diversification has led tocombined mining and agricultural pursuits. See Banchirigah (2008) andMaconachie and Binns (2007).

4. However, a dichotomy exists between legislation and implementationpertaining to small-scale mining. Thus, while the mining legislation hasled to a significant influx of investments by large-scale miningcompanies, the vast majority of small-scale miners still engage ininformal mining activities, unaware of the actual content of thelegislation (Fisher, 2007).

5. Notably in Chunya, Geita, Kahama districts. Most of these companiesdo not prospect actively, but continuously renew their licenses, anticipat-ing a future profitable resell if gold is discovered (Jønsson & Fold, 2009;Lange, 2008).

6. Our field study encompassed an age-stratified random sample of 45miners and nine former miners in each of the settlements, totaling 108respondents of whom nine provided additional oral histories (anadditional 41 claim owners and pit holders were interviewed in connectionwith the study). Survey sampling took place at off-work recreational siteswhere miners gathered (a board game site, a TV saloon, two bars, twolocal brew hang-outs, and a coffee salon). Each site was divided into a3 � 3 cell matrix and given a number to facilitate a random draw ofpersons of the targeted age groups. The 18 former miners were selectedpurposively to gain a cross-section of respondents of various ages andincomes. Our sampling procedures were biased toward successful miners

given that the survey took place in a mining site, omitting those who mayhave quit mining due to economic ruin, social demoralization, poorhealth, or other reasons. Worksite interviews had to be avoided given theobstruction to work and lack of privacy they posed.

7. Along with daily chance encounters with miners as site observers, atleast an hour a day was spent chatting at the local coffee shop in Londoni,a meeting place for pit holders and claim owners. Moreover, at the time ofthe survey, participatory research and technical workshops on mercuryusage in mining with more than 200 participants had been conducted for 5months in each settlement (Jønsson, Appel, & Chibunda, 2009).

8. In 2006–08, 1 US$ was equivalent to between Tsh 1,050 and 1,250.

9. There may, however, have been some sampling bias against claimowners who tend to be more private figures than pit holders and not aslikely to socialize at public recreational sites. Only one claim owner fromLondoni and three from Matundasi appeared in our random sample.

10. Shifts usually amount to a period of anywhere between one and fourweeks determined by the pit holder in consultation with the claim ownerwith regard to the gold yield following the sequence of excavation worknecessary to construct galleries, blast gold-bearing reefs, and hoist rocks tothe ground. A shift depends on the output of the excavation and themorale of the workers roughly determined by the nature of goldproduction. Some pit holders set production targets and stop afterachieving these.

11. These included small businesses: running bars, restaurants, guesthouses, hair salons, an occasional disco, a butchery, a general shop;services such as machine maintenance, renting out bikes, shoe repair, shoe-shining, ironing clothes in a tailor’s shop, carrying firewood for people;and self-employment in maize trading, house construction, and as acarpenter of local beds; as well as employment as a bus conductor and in aclothes shop.

REFERENCES

Banchirigah, S. M. (2008). Challenges with eradicating illegal mining inGhana: A perspective from the grassroots. Resources Policy, 33(1),29–38.

Barret, C. B., Reardon, T., & Webb, P. (2001). Nonfarm incomediversification and household strategies in rural Africa: Concepts,dynamics and policy implications. Food Policy, 26, 315–331.

Bryceson, D. F. (1996). De-agrarianization and rural employment in sub-Saharan Africa: A sectoral perspective. World Development, 4(1),97–111.

Bryceson, D. F. (1999). African rural labor, income diversification andlivelihood approaches: A long-term development perspective. Reviewof African Political Economy, 80, 171–189.

Bryceson, D. F. (2000). Peasant theories and smallholder policies: Pastand present. In D. F. Bryceson, C. Kay, & J. Mooij (Eds.),Disappearing peasantries? Rural labor in Africa, Asia and Latin America(pp. 1–36). London: Intermediate Technology Publications.

Bryceson, D. F. (2002a). The scramble in Africa: Reorienting rurallivelihoods. World Development, 30(5), 725–739.

392 WORLD DEVELOPMENT

Bryceson, D. F. (2002b). Multiplex livelihoods in rural Africa: Recastingthe terms and conditions of gainful employment. Journal of ModernAfrican Studies, 40(1), 1–28.

Bryceson, D. F., & Jamal, V. (1997). Farewell to farms: Deagrarianisationand employment in Africa. Aldershot, UK: Ashgate Publishing Ltd..

Bryceson, D. F., & Mwaipopo, R. (2009). Rural–urban transitions inNorthwestern Tanzania’s mining frontier. In J. Agergaard, N. Fold, &K. Gough (Eds.), Rural urban dynamics: Livelihoods, mobility andmarkets in African and Asian Frontiers. London: Routledge.

Butler, P. (2004). Tanzania: Liberalisation of investment and the miningsector. In B. Campbell (Ed.), Regulating mining in Africa: For whosebenefits? (pp. 67–80). Uppsala: Institute for African Studies.

Chachage, C. S. L. (1995). The meek shall inherit the earth but not themining rights: Mining and accumulation in Tanzania. In P. Gibbon(Ed.), Liberalised development in Tanzania (pp. 37–108). Uppsala:Institute for African Studies.

Drechsler, B. (2001). Small-scale mining and sustainable development withinthe SADC Region. London: International Institute of Environmentand Development/World Business Council for Sustainable Develop-ment.

Ellis, F. (2000). Rural livelihoods and diversity in developing countries.Oxford: Oxford University Press.

Ellis, F., & Freeman, H. A. (Eds.) (2005). Rural livelihoods and povertyreduction policies. London: Routledge.

Emel, J., & Huber, M. T. (2008). A risky business: Mining, rent and theneoliberalization of ‘‘risk”. Geoforum, 39(3), 1393–1407.

Fisher, E. (2007). Occupying the margins: Labor integration and socialexclusion in artisanal mining in Tanzania. Development and Change,38(4), 735–760.

Francis, E. (2000). Making a living: Changing livelihoods in rural Africa.London: Routledge.

Godoy, R. (1985). Mining: Anthropological perspectives. Annual Reviewof Anthropology, 14, 199–217.

Grawert, E. (1998). Making a living in rural Sudan. London: MacmillanPress Ltd..

Hilson, G. (Ed.) (2003). The socio-economic impacts of artisanal and small-scale mining in developing countries. The Netherlands: A.A. Balkema.

Hilson, G. (2007). What is wrong with the global support facility forsmall-scale mining?. Progress in Development Studies, 7(3), 235–249.

Hilson, G., & Potter, C. (2005). Structural adjustment and subsistenceindustry: Artisanal gold mining in Ghana. Development and Change,36(1), 103–131.

Hilson, G., Yakovleva, N., & Banchirigah, S. M. (2007). ‘To move or notto move’: Reflections on the resettlement of artisanal miners in theWestern Region of Ghana. African Affairs, 106(424), 413–436.

Homewood, K. (Ed.) (2008). Rural resources and local livelihoods in Africa.Oxford: James Currey.

ILO (1999). Social and labour issues in small-scale mines: Report fordiscussion at the tripartite meeting on social and labour issues in small-scale mines. Geneva: International Labour Organization.

Ip, C. Y. (2008). The end of careers? Changes of male employment careers inBritain between 1955 and 2004. DPhil thesis, University of Oxford.

Jønsson, J. B., Appel, P. W. U., & Chibunda, R. (2009). A matter ofapproach: The retort’s potential to reduce mercury consumptionwithin small-scale gold mining settlements in Tanzania. Journal ofCleaner Production, 17(1), 77–86.

Jønsson, J. B., & Bryceson, D. F. (2009). Rushing for gold: Mobility andsmall-scale mining in East Africa. Development and Change, 40(2),249–279.

Jønsson, J. B., & Fold, N. (2009). Handling uncertainty: Policy andorganizational practices in Tanzania’s small-scale gold mining sector.Natural Resources Forum, 33(3), 211–220.

Kanter, R. M. (1989). When giants learn to dance. Mastering the challengeof strategy, management and careers in the 1990s. New York: Simonand Schuster.

Konings, P. (in press). African workers’ trade union identity in the wakeof structural adjustment in Cameroon. In D. F. Bryceson (Ed.), HowAfrica works: Occupational change, identity and morality. London:Practical Action Publishing.

Lange, S. (2006). Benefit streams from mining in Tanzania: Case studiesfrom Geita and Mererani. Bergen: Chr. Michelsen Institute.

Lange, S. (2008). Land tenure and mining in Tanzania. Bergen: Chr.Michelsen Institute.

Macdonald, C., & Roe, R. (2007). Tanzania country case study. Thechallenge of mineral wealth: using resource endowment to fostersustainable development. International Council on Mining & Metals(ICMM). 26.03.08. <http://www.icmm.com/publications/1700RECC-STanzaniaFINAL.pdf>.

Maconachie, R., & Binns, T. (2007). ‘Farming miners’ or ‘miningfarmers’? Diamond mining and rural development in post-conflictSierra Leone. Journal of Rural Studies, 23(3), 367–380.

Madulu, N. F. (1998). Changing lifestyles in farming societies of Sukuma-land: Kwimba District, Tanzania (Vol. 27). Working paper. Leiden:Africa Studies Centre, University of Leiden.

Meagher, K. (in press). Social capital or social exclusion? Social networksand informal manufacturing in Nigeria. In D. F. Bryceson (Ed.), HowAfrica works: Occupational change, identity and morality. London:Practical Action Publishing.

Meagher, K. (2001). The bargain sector: Economic restructuring and thenon-farm sector in the Nigerian savanna. Aldershot, UK: AshgatePublishing Ltd..

Meagher, K., & Mustapha, R. (1999). Not by farming alone: The role ofnon-farm incomes in rural Hausaland. In D. F. Bryceson, & V. Jamal(Eds.), Farewell to farms: De-agrarianisation and employment in Africa(pp. 63–84). Aldershot, UK: Ashgate.

Mwaipopo, R., Mutagwaba, W., Nyanga, D., & Fisher, E. (2004).Increasing the contribution of artisanal and small-scale mining to povertyreduction in Tanzania. London: Department for International Devel-opment.

Phillips, L. C., Semboja, H., Shukla, G. P., Sezinga, R., Mutagwaba, W.,Mchwampaka, B., et al. (2001). Tanzania’s precious mineral boom:Issues in mining and marketing. Washington, DC: USAID.

Roberts, A. D. (1986). The gold boom of the 1930s in Eastern Africa.Journal of the Royal African Society, 85(341), 545–562.

Tanzania, United Republic (1998). The mining act. Ministry of Energy andMinerals, United Republic of Tanzania. Dar es Salaam: GovernmentPrinter.

Tanzania, United Republic (1999). The mining regulations. Ministry ofEnergy and Minerals, United Republic of Tanzania. Dar es Salaam:Government Printer.

Tanzania, United Republic (2005). National strategy for growth andreduction of poverty (NSGRP). Vice President’s Office. Dar es Salaam:Government Printer.

Tanzania (2003). Population and housing census 2002. General Report. Dares Salaam: National Bureau of Statistics.

Available online at www.sciencedirect.com