Introduction^{Footnote 1}

This article analyzes trends in demography and labor relations in Tanzania between 1800 and 2000. The main aim is to disclose the changes in population size and composition and connect them to continuities and shifts in labor relations. In this paper, I will discuss the quality and nature of the sources available on demography and labor relations in Tanzania for four cross-sections (1800–1900–1950–2000). When necessary, I will refer to the methodological paper in which I present and discuss the calculation methods that I have used to reach the data for the database.^{Footnote 2} I do not claim that the figures that I have obtained are accurate; they are subject to large margins of error because the basis for estimations is often in itself very poor. Nonetheless, they can provide a degree of magnitude, which is useful for the goals of the Collaboratory and Global Labor History in general.

The dataset also includes information on Zanzibar. Even if not extremely relevant from a statistical point of view, Zanzibar is nonetheless critical in the analysis of labor relations, owing to the peculiarities of its labor market, as well as the better availability of sources.^{Footnote 3}

Shifts and continuities in labor relations are clearly connected to the historical trajectories of the area now known as the United Republic of Tanzania. Therefore, the first section of this paper will present a brief historical overview of the country and particularly of the events that produced significant changes in demography and labor relations. The second section discusses the quality and nature of the sources available on demography and labor relations in Tanzania for the given cross-sections. In light of the available sources, the third section analyzes the main trends in the population size and composition. The final section provides a picture of the major shifts in labor relations, their causes and consequences, in connection with the major historical events that took place in Tanzania over the last two centuries.

Brief Historical Background of the Country from 1800

At present, Tanzania is a country with a population of 44,928,923 occupying an area of almost 950,000 square kilometers.^{Footnote 4} During the scramble for Africa, the area was occupied by Germany, which combined present-day mainland Tanzania, Rwanda and Burundi to create the colony of German East Africa (GEA). In 1890, Great Britain established a protectorate over Zanzibar. After Germany lost the First World War, the area was transferred to Belgium (Rwanda and Burundi) and Great Britain (Tanganyika Territory) through a League of Nations mandate. In 1961, Tanganyika reached independence and in 1964, following the union with Zanzibar, it became the United Republic of Tanzania.

In 1800, the coastal areas of Tanzania were part of the Indian Ocean commercial world. Thanks to the monsoon winds – the driving force of the Indian Ocean commerce – India and the Arabian Peninsula had developed close commercial relationships with East Africa.^{Footnote 5} In contrast, the interior regions of Tanzania were characterized by a wide network of interregional trade and commercial contacts with the coast were limited. The nineteenth-century demand for ivory in order to produce luxury items for the growing Western middle class and the demand for slave laborers from Zanzibar plantations were the main causes for integrating the interior regions into the commercial empire of Zanzibar. Coastal traders were able to penetrate inland thanks to the Omani state, which supported their commercial enterprises, the Indian financers on the coast, who provided long-term credit facilities, and to the extremely favorable terms of trade that African commodities, particularly ivory, enjoyed throughout the nineteenth century.^{Footnote 6} At the same time, a growing awareness of the greater value of ivory led enterprising African businessmen from the interior regions to increase regional trading despite hazards and risks of the long journey to the coast.^{Footnote 7} The development of long-distance trade with the coast gave birth to a new class of paid laborers, namely porters. Porters represent a peculiar case of an early development of wage labor in pre-colonial East Africa.^{Footnote 8} As the century progressed, a steady increase in the price of ivory expanded the area in which Swahili, Arab and African traders operated.^{Footnote 9} This led to the development of urban centers along the most important caravan routes, such as Tabora, Ujiji and Bagamoyo, whereas well-developed urban centers already existed on the coast.^{Footnote 10}

The first years of the German colonial period were characterized by a continuation of the economic patterns that had originated in the nineteenth century. Ivory, later replaced by wild rubber, remained the main export item and trade continued to center around Zanzibar. Caravan trade and the number of porters traveling along the caravan routes remained stable between the 1890s and early 1900s.^{Footnote 11} With the construction of the colonial railways and the introduction of new export crops, new economic patterns started to develop. In 1905, the Tanga line was completed and by 1914, the central railway reached Lake Tanganyika. Railway construction laborers were brought from India, as well as recruited locally, while forced labor was also commonly used.^{Footnote 12} In the 1890s, coffee was introduced by German settlers. Since the beginning of the twentieth century, the number of German settlers increased and they started to experiment with new crops, especially cotton, sisal and rubber. In the last decade of German rule, the productivity of the settlers’ plantations increased along with a remarkable rise in investments. This development was interrupted by the outbreak of the First World War. Africans were recruited as porters and soldiers and German settlers also took part in the war. When Germany lost the war, the British took over the area and named it Tanganyika Territory.

Between the wars, two processes characterized Tanganyika’s economic history: an increase in the area’s involvement in the global economy and a process of regional differentiation.^{Footnote 13} Integration into the global economy was favored by the expansion of new areas for cash crop production – especially coffee, cotton and tobacco – and the development of roads and motor transport. This was accompanied by the further development of African farming, especially coffee by the Chagga. In 1929, the development of the export economy – which had begun in the 1920s – was interrupted by the Great Depression. This reduced the price of Tanganyika’s cash crops on the international markets. In the 1930s, the plantation economy gradually started to recover and became more and more based on the cultivation of sisal. However, Britain failed to create a real structural economic change and basically favored a process of regional differentiation and intensification of labor migration to cash-producing areas.^{Footnote 14}

The outbreak of the Second World War led to a return to widespread forced and conscript labor. During the war, a total of 74,000 forced laborers were enrolled by private employers for the production of sisal, pyrethrum, rubber, rice and essential foodstuffs for troops and prisoners of war.^{Footnote 15} A total of 87,000 soldiers volunteered or were conscripted for the war.^{Footnote 16} After the end of the war, large-scale agricultural projects were implemented in Tanganyika. One such project was the East African Groundnut Scheme in the Southern Province, a project conceived to address the world shortage of vegetable oils.^{Footnote 17} The East African Groundnut Scheme was a failure, but testified to Great Britain’s commitment towards the economic development of its colonies in the post-war period.

New settlement patterns also developed in this period and the colonial authorities started to show some concern towards the increasing urban population of the East African colonies. According to the East Africa Royal Commission Report, “Settlements of closely-packed African huts are to be seen on the fringes of all the larger towns in East Africa” and “Tanganyika is faced, equally with the other territories, with the problem of establishing reasonable conditions of life of its urban populations.”^{Footnote 18} In 1950, the urban population was still limited (between 1.5 and 2 percent), but it is evident that a process of urbanization was set in motion during this cross-section, which started to transform labor relations in the urban areas. African laborers in the main towns increased, working not only as unskilled laborers, but also as clerks, foremen and artisans, businessmen and builders.^{Footnote 19} The presence of these laborers was often temporary. They generally retained their landholdings in the areas where they came from and women and children were left behind to cultivate the land.^{Footnote 20}

In December 1961, Tanganyika obtained independence from Great Britain and in 1964, after the union with Zanzibar, it became the United Republic of Tanzania. The most important turning point after independence was the 1967 Arusha Declaration. With this declaration, President Julius K. Nyerere introduced the basic elements of Ujamaa: Tanzanian socialism based on self-reliance and a strong emphasis on agricultural development and production. The objective of Ujamaa was to create a new Tanzanian economy, independent from foreign aid and devoted to the foundation of a more equal society. The economic means to reach the objectives of the Arusha Declaration were the production of Tanzanian cash crops, such as sisal, cotton and pyrethrum, and of food crops, such as maize and wheat.^{Footnote 21} The project was based on the transfer of peasants to communal villages, where wage labor was banned and means of production were commonly owned. In 1973, almost two million Tanzanians were living in communal villages and, after the forced villagization of 1976, 85 percent of the total rural population lived in communal villages.^{Footnote 22} From an economic point of view, the results of Ujamaa were poor, partly owing to the low prices of Tanzanian cash crops on the international markets. By the end of the 1970s, Tanzania started to accumulate a rising international debt. Subsequently, in 1986, President Ali Hassan Mwinyi signed the first Structural Adjustment Programme with the International Monetary Fund. In the 1980s, GDP rates decreased significantly together with agricultural and industrial production.^{Footnote 23}

As we are going to see in the following sections, agriculture remained the main occupation of Tanzanian population throughout the nineteenth and twentieth century, with specific developments of commodified labor in some sectors.

“Huts are Easier to Count than Heads:”^{Footnote 24}Counts, Enumerations and Censuses

During the various Africa workshops of the Collaboratory, we discussed the methodological problems related to demography in Africa, especially before 1950.^{Footnote 25} One of the most problematic issues is the almost total absence of reliable information on the 1800 cross-section. With regard to Tanzania, some fragmentary information can be obtained from estimates made by travelers and merchants who visited the coast. However, no information is available on the interior regions. With the start of the colonial period, the sources gradually become more abundant, but their reliability is hindered by the attitude to enumerations.^{Footnote 26} Early colonial population surveys were conducted primarily for tax purposes and to fix quotas for forced labor, which were “both excellent motives for avoiding enumeration.”^{Footnote 27} Therefore, they almost invariably resulted in underestimations of the population.

Generally, the number of children was also underestimated.^{Footnote 28} In many parts of East Africa, counting children was considered bad luck. As Goldthorpe found out when interviewing his students from various parts of East Africa, “there is a general belief that counting children makes God think that one has too many of them and so He relieves him of some of them.”^{Footnote 29} Colonial representations of population were in themselves part of a larger discourse of power, control and the exercise of authority, but their results were clearly limited by the reach of the colonial state.^{Footnote 30} Nonetheless, an analysis of the characteristics and the purposes of the various enumerations and censuses provides an understanding of the mechanisms behind them and can therefore reveal the demographic processes they purport to measure.^{Footnote 31}

In the German colonial period, population estimates were entirely based on tax returns as the administrative capacity of the German colonial state was too limited to carry out general surveys of the population.^{Footnote 32} Colonial estimates were based on implicit assumptions.^{Footnote 33} The size of the total population was calculated based on the number of male taxpayers multiplied by 3.5, which was assumed to represent the number of dependents (1 woman and 2.5 children) of every male taxpayer.^{Footnote 34} Later censuses conducted by the British showed that the German estimations for GEA were even worse than those that the British had conducted during the same period for Uganda and Kenya.^{Footnote 35}

With the transfer of power and the increased capacity of the colonial state, more frequent enumerations and counts were conducted, which resulted in detailed demographic studies.^{Footnote 36} The British carried out their first enumeration in 1921, another one in 1929 and a census in 1931. These population surveys were again based on implicit assumptions: adult men made up 49 percent of the total population and the number of children under the age of 15 was estimated at 37 percent of the total adult population. As taxes were supposed to be paid by all adult men, the number of women and children was calculated accordingly. Only in 1948, did the British carry out the first real census in the East African Territories, which was the first hut-to-hut enumeration of the population of Kenya, Uganda, Tanganyika Territory and Zanzibar Protectorate. This was a well-conceived census, carefully organized for two years by the East African Statistical Department and its director, C.J. Martin, who planned it “as a military exercise.”^{Footnote 37} The census was completed within a few days – most of the enumerators completed their work in four days – and this helped to reduce common mistakes like double counting.^{Footnote 38} This is the first census that provides us with reliable information on the age composition of the population. Enumerators were trained to estimate people’s age by reference to historical events that had happened in their districts, like the Maji-Maji war, the arrival of the railway, or the recruitment of soldiers for the First World War. In this way it was possible to link age groups to historical events.

Demographic studies were often connected to labor surveys, and colonial authorities perceived and counted the population in terms of labor power.^{Footnote 39} One of the main concerns of the colonial state was “the perpetual search for labor,”^{Footnote 40} and the results of censuses and enumerations were interpreted to explain labor shortages, which were often blamed on low fertility rates or poor maternal care. In 1911, the Germans carried out their first detailed labor survey. The British only started to carry out detailed labor surveys towards the end of the colonial period, thanks to the work of the East African Statistical Department.^{Footnote 41} In 1949, a Labor Census was carried out; followed in 1950, by an analysis of The Pattern of Income Expenditure and Consumption of African Laborers in Dar es Salaam and a Report on the Analysis of the Sample Census of African Agriculture; and in 1952, an Enumeration of African Employees. These sources have clear limits, and they often provide contradictory results. Nonetheless, they provide information on the structure of the labor market and a context for highlighting shifts in labor relations.

Since the 1960s, our knowledge of the demography of the African continent has increased considerably, mostly thanks to the fact that the United Nations Commission for Africa has carried out systematic overviews of the African population.^{Footnote 42} Demographic data for the 2000 cross-section have been obtained from the 2002 Population and Housing Census, whereas the analysis of labor relations is based on the Integrated Labor Force Survey 2000/2001 Analytical Report and the 2001 Child Labor in Tanzania Country Report.^{Footnote 43} These data are questionable, owing to invariable problems like double counting, but they have been considered reasonably reliable for the general purpose of the database.^{Footnote 44}

Given the limits of the sources highlighted so far, the next section aims at portraying a general picture of the main shifts in the population size and composition of the area now known as the United Republic of Tanzania.

The Composition of Tanzania’s Population (1800–1900–1950–2000)

Little information is available on the size and growth of Tanzanian population in the nineteenth century. According to both Iliffe and Kjekshus, the population of East Africa was relatively stable or expanded slowly throughout the nineteenth century. Kjekshus calls into question the colonizers’ portrayal of East Africa as a violent victim of the slave trade and internecine wars. He uses economic evidence to support his hypothesis of a stable population or slow population growth in the nineteenth century, only interrupted in the 1890s by colonial occupation and wars.^{Footnote 45} Iliffe sees in the introduction of new crops, like maize and rice, evidence of an agricultural change that supported a population growth in the nineteenth century.^{Footnote 46} In contrast, Hartwig is of the opinion that the existing fragments of evidence are an indication of high mortality as a result of warfare and slavery as well as epidemics, such as smallpox and cholera, famines – one very big one in the 1830s – and several, widespread droughts.^{Footnote 47} Along the same lines, Koponen points out that there were some regions that witnessed a population growth and others a demographic decline, but the general trend was a disruption of population patterns, culminating in the famine of the 1890s.^{Footnote 48} Manning assumes that death rates declined from the mid-nineteenth to the mid-twentieth century, whereas birth rates remained more or less stable, so that net rates of population growth increased over time on the continent. He then provides a list of social circumstances (for instance, slave trade, migration of free people, famine, and epidemics) and calculates their possible impact on African societies. He concludes that slave trade regions like Tanzania witnessed a population decline at least until the 1880s.^{Footnote 49} According to Lovejoy, the number of slaves traded in East Africa during the nineteenth century was almost 1,650,000, 47.5 percent of which were retained in the area, being employed in the East African plantations along the coast.^{Footnote 50} What is today Tanzania was therefore not only a slave-exporting, but also a slave-importing area. Slaves, for instance, were imported to replace caravan porters while they were on journey, or were purchased by traders living in the urban centers of the interior and employed in domestic or plantation work.^{Footnote 51}

From the 1890s there was a significant population decline, caused by an epidemic of rinderpest (1891–1892) and by colonial wars. The rinderpest epidemic spread southwards from the Horn of Africa. Between 1897 and 1898, East Africa fell victim to a widespread famine, which included Tanganyika. We do not have enough data to affirm with any certainty how widespread and destructive these events were. Given that Iliffe and Kjekshus use wider categories of analysis, like settlement patterns and agricultural change, in order to analyze demographic trends for the whole area, I have based my estimates on their hypothesis of a stable or slowly growing population during the nineteenth century. Moreover, I have considered the disruptive effects of the slave trade, as suggested by Manning, and the fact that many slaves were imported and retained locally, as highlighted by Lovejoy. On the basis of the demography of the area in the early colonial period (about 4,000,000 people at the beginning of the twentieth century, see below) and taking into consideration the general nineteenth-century historical trends, the total population of the 1800 cross-section might have ranged between 3,500,000 and 4,000,000. It is not possible to know the composition of the population for the 1800 cross-section, as more precise studies are needed on how the epidemics, famines and wars of the last decade of the nineteenth century affected women and children.

Given that German estimates were based on tax returns, my starting point for estimating the total population for the 1900 cross-section is with the data available after 1905, when a poll tax replaced the hut tax in many districts.^{Footnote 52} Estimations based on the number of adult men paying tax rather than on the number of huts undoubtedly produce rough results, but at least these are more accurate. According to estimates, in 1906 the total population was 4,009,500.^{Footnote 53} If considered in relation to the 1948 census data, this figure could be plausible only when considering a constant population increase. However, this was not the case, given that some events that occurred during the German colonial period had disastrous effects on the population size. According to estimates, repressing the Maji Maji revolt (1905–1907) – an uprising in the south of the colony against the forced cultivation of cotton – resulted in 250,000–300,000 mortal victims.^{Footnote 54} During the First World War, German East Africa was a battlefield that killed civilians, porters and soldiers.^{Footnote 55} At the end of the war, returning porters and soldiers carried with them the Spanish influenza, an epidemic that actually killed more people than the war itself.^{Footnote 56} Owing to the combination of these events, there is reason to believe that the population of Tanzania decreased between 1900 and 1925.^{Footnote 57} Contrary to the general trend of colonial demographic surveys, which tended to underestimate the population, it seems that the Germans overestimated the total population of GEA. Most likely, this overestimate was due to the fact that there were fewer dependents per male tax payer than what the Germans had assumed. Culwick and Culwick studied fertility in the Ulanga District, southern Tanzania, and found that in the 1930s the net reproduction rate was 0.73. They documented a decline in the population from the beginning of the twentieth century. This was corroborated by maps of Ulanga District during the early colonial period, which showed that there were more villages in that period than in the end of the 1930s. This area, which was one of the most fertile areas of East Africa, was also involved in the Maji Maji revolt. Culwick and Culwick therefore concluded that for the period of 1905–1935, similar trends might possibly apply to areas outside of Ulanga district.^{Footnote 58} Actually, the 1921 census, despite all of its limits, revealed that there were fewer children than assumed. In coastal areas, the adult–children ratio was 1 to 2, whereas in other districts it was 1 to 1.5. In the district of Ujiji, where the Germans had assumed 5 dependents per taxpayer, the actual ratio turned out to be 1 adult per 3 dependents.^{Footnote 59}

Therefore, for the 1900 cross-section I have estimated that the population of German East Africa, excluding Rwanda and Burundi, was between 3,500,000 and 3,800,000, to which I have added the population of Zanzibar, c. 200,000.^{Footnote 60}

Regarding the composition of the population, it is almost impossible to establish how many children and women lived in the colony during this cross-section. According to a German estimate, in 1906 the number of children between 0 and 15 years was 42 percent of the total population. This figure is definitely too high, for the reasons highlighted above, and taking into consideration the overall high child mortality of GEA.^{Footnote 61} Although this is questionable, I have postulated that the number of children between 0 and 15 years could have been around 35 percent of the population, and from this figure I have then calculated the number of children between 0 and 5 years. No specific data are available on elderly people, only estimates on life expectancy in the African continent. In the early nineteenth century, this was between 20 to 25 years and rose to 35 by 1950.^{Footnote 62}

The Germans estimated the number of women according to the number of taxpayers. In 1931, the British tried for the first time to enumerate the total population of Tanganyika with the help of the native authorities. They estimated that 53 percent of the total population consisted of women.^{Footnote 63} This data was confirmed by the 1948 census, which estimated that the portion of women was 52 percent. Every enumeration or census conducted from the 1930s onwards showed that women comprised slightly more than 50 percent of the total population. We do not have reliable information on female mortality rates nor evidence of a higher female mortality as a result of the Maji Maji war, the epidemics of the early twentieth century and the First World War. Perhaps even the available evidence points towards the opposite, given that soldiers and porters were almost all men. Therefore, I have guesstimated that in this cross-section, women comprised slightly more than half of the population. This is of course debatable, but 51–53 percent seems to be a plausible rate.

For the following cross-section, namely 1950, I have used data from the 1948 census. The last estimate that still used the old methods estimated the total population at 5,838,888 in 1947, whereas the 1948 census gave the total population at 7,408,000. This discrepancy is mostly the result of there being more children than initially assumed.^{Footnote 64} Analysis based on later censuses stated that the underestimation of the total population in 1948 was between 1.9 and 3.6 percent, which adds up to a total population of about 7,600,000.^{Footnote 65} When adding to this the total population of Zanzibar and Pemba (264,162), we get a total population of about 7,900,000.^{Footnote 66} In the Tanganyika Territory, women comprised 52 percent of the total population, whereas in Zanzibar they comprised 47.5 percent.^{Footnote 67}

Before the 1948 census, two age groups were considered in the various colonial enumerations: 0–15 years old (i.e., non-taxpayers) and over 15 years old (i.e., taxpayers).^{Footnote 68} Consequently, a distinction was made between adults and non-adults. This was determined according to a vaguely-defined category of “physical development” or “sexual maturity,” which local authorities that carried out the census were asked to determine.^{Footnote 69} In practice, the way to establish if a boy had reached the age of 15 was to look under his arm and see whether he had any underarm hair.^{Footnote 70} The 1948 census was the first census to provide fairly reliable information on age groups. As mentioned above, enumerators were trained to estimate people’s age by referring to historical events, like wars or the opening of the railway. This allowed them to link age groups to historical events. The age groups in the census form were 0–1, 1–5, 6–15, 16–45 and over 46.^{Footnote 71} The age groups clearly got mixed up a bit, and it is very likely that many girls between the ages of 14 and 15 were included in the 16–45 age group and that 16 years old boys were included in the 6–15 age group. Nevertheless, we at least have a more reliable picture of the age composition of the Tanganyika population. Almost 19 percent of the total population were children between the age of 0 and 5, to which I have added the number of children in Zanzibar.^{Footnote 72} According to C.J. Martin, the director of the East African Statistical Department, men belonging to the group over the age of 46 were too old to work and were, in many cases, exempted from paying poll tax. In Zanzibar, approximately 9 percent of the total population consisted of people over the age of 60.

Finally, for the 2000 cross-section we have a total population of 34,443,603, which includes Zanzibar, of which 51.13 percent were women.^{Footnote 73} The children age groups are 0–4 years old and 5–9 years old. The number of 0–5 year old children has been calculated accordingly, and represents about 19 percent of the total population. People over 65 represent 3.9 percent of the total population.

In 1800, the area that corresponds to present-day Tanzania probably had a population of between 3,500,000 and 4,000,000 people, which was 3,700,000–4,000,000 in 1900, 7,900,000 in 1950 and almost 35,000,000 in 2000. Not surprisingly, the population doubled from 1900 to 1950, and increased more than fourfold from 1950 to 2000. As discussed at length above, some data, especially on the composition of the population, can only be roughly estimated or even guesstimated. Nonetheless, having a broad, even if sometimes vague, idea of these figures contributes to the aim of the project in obtaining a picture of the active and inactive population for the various cross-sections, which in turn provides a context for analyzing the major shifts in labor relations.

Describing and Quantifying Shifts in Labor Relations