Version classiqueVersion mobile
OpenEdition Books Africae Africae Studies Indian Africa East African Indians: How Many Ar...

Indian Africa

 | 
Adam Michel

East African Indians: How Many Are They?

Laurent Nowik

Texte intégral

Introduction

  • 1 The author sincerely thanks Kenya’s Central Bureau of Statistics and Uganda’s Bureau of Statistics (...)
  • 2 Ugandan 1969 Census.

1In East Africa (Kenya, Tanzania and Uganda), the communities of Indian and Pakistani origin went through difficult times during the initial years of independence in newly established states which adopted discriminatory policies against them, and which in the case of Uganda even led to their expulsion1. These events reversed the previous demographic dynamics characterised by uninterrupted growth. Whereas in 1969 the population of the Indian diaspora in Uganda numbered 75,000, a third of whom being Ugandan nationals2, only a few hundred Indians, all of them probably Ugandan citizens, remained in the country after the events of 1972. In Kenya, the Indian population declined by about 100,000 between 1962 and 1979, therefore being roughly halved. In Tanzania, 20,000 Indians were estimated to have left Tanzania following the nationalization policy adopted by the socialist government (Prunier 1990: 239).

2It is known that even though the establishment of Indian settlements in East Africa started several centuries ago, massive immigration does not date back to before the arrival of the British. The fact remains that a number of people of Indian origin today have offspring born in Africa over more than five generations and no longer have links to the Indian subcontinent, though they have not undergone a complete process of acculturation into the African society.

3Even though most of the Indians living in East Africa now hold the nationality of their country of residence henceforth (at least in Kenya and Tanzania) and have known no other home, their counting has always caused reticence, either over concerns of not isolating these people from other citizens or over fears of underlining their demographic importance, in view of the fact that their economic strength gives rise to disputes. In this respect, policies on demographic data collection and the dissemination of census results can be seen, in each of the three East African countries, as indicators of the differential treatment of ethnic diversity.

  • 3 Whereas the terms “race”, “racial group”, “tribe” and “ethnic group” were more or less synonymous d (...)

4Concerned about national unity, Tanzania has, since its independence, stopped making any reference to the ethnic variable in censuses, whether of Africans or non-natives3. The Tanzanian government has also encouraged naturalization of Indians, who have indeed become “ordinary” citizens in public statistics because they are no longer categorized statistically. There is consequently no reason to carry demographical studies on the Indians of Tanzania any longer, except from a historical perspective. However, in socio-political reality, this does not mean that citizens of Indian origin have become citizens “like the others”, or that, within their own community, they would not be concerned about counting themselves.

5Uganda, the country in which the Indian question is still the most difficult to talk about, has maintained the principle of statistical categorisation of ethnic origin, but has, since the events of 1972, desisted from releasing such information to the public. Up to 1969, publications of censuses followed the current practices in place under the British protectorate and distinguished between “African” and “non-African”. About the second group, special booklets on “non-native” citizens specified the nationality, ethnicity and geographical origin of “Europeans”, “Arabs” and “Asians”. After 1969, public statistics simply distinguishing Ugandans and non-Ugandans, not mentioning any other criteria of distinction. Thus in the 1991 census – which mentioned ethnic origin in the forms distributed to the public – no information about ethnicity was published, even ignoring the existence of citizens of Indian origin as well as the percentage of Indians in the category of non-Ugandans. By doing so, Uganda came close to Tanzania’s position in terms of the release of statistics, making the decision not to distinguish citizens according to their origin in order to assuage cleavages and strengthen national unity, after several decades of bloody conflicts.

6Following the 2001 census, the ethnic variable resurfaced and in 2005 the number of the nine largest “native” ethnic groups was released to the public. As for foreigners, only “Indians” were specially mentioned without, however, giving any indication of the total number of Indian or Pakistani nationals, as we will see further on.

7Although it was not published, statistical data on the origin of Indian residents had been collected, as indicated above, during the two censuses of 1991 and 2001. Following our diligent requests, the Uganda Bureau of Statistics (UBOS) agreed to provide the information we sought on the 1991 census. For the 2001 census, the UBOS only provided spreadsheets showing the number of residents who had indicated they were of “Asian” nationality or origin (27 % of them being Ugandan-born). These spreadsheets do not provide any answer to the question of how many Ugandan citizens are of Indian origin, neither to how many people from the Indian sub-continent hold another nationality (African or British). The available information based upon the data provided to us leads to underest imations of the number of Indians living in Uganda using multiple passports. On the other hand, we will be able to present the main socio-demographic characteristics of these populations.

8During various censuses conducted since the end of British colonial rule, the Kenyan government has regularly published statistics on the ethnic origin of its people, which consequently included the Indian population, as well as on citizenship. During the last census in 1999, the Central Bureau of Statistics (CBS) decided not to release information on ethnicity and citizenship, even though it was part of the data collected by census takers. Upon our request, the CBS agreed to release to us, for strictly scientific purposes, the set of data on Indians collected in the last two censuses (1989 and 1999). Of the three countries studies, the most precise demographic information will therefore concern Kenya.

9The following three sections on each of the three countries studied follow an order chosen by the author, and are hardly interdependent. The reader who wants to learn about one country only may therefore read only the section of interest.

Indian Tanzanians

10As mentioned above, since the time of independence, Tanzanian public statistics no longer numerate African ethnic groups (120 in number, according to censuses carried out in the colonial era) or citizens from other continents. The approach adopted in this section is therefore essentially historical, bearing in mind that the most detailed data date back to the period between 1931 and 1967. Indian population estimates during the German colonial period (1891–1920) are incomplete and sometimes contradictory, since the first counting carried out in 1913 cannot be considered to be a real census. From 1921, censuses commissioned by the British administration provided regular information on the population, with the exception of the interval preceding and following the Second World War (1931–1948).

11The presentation will be grounded upon seven censuses:

  • British censuses conducted in 1921, 1928 and 1931;

  • the United Nations population report for Tanganyika (1st September
    1949) written before the release of results of the 1948 census;

  • the 1948 census;

  • inter-censal estimates by the East African Statistical Department released in 1952;

  • the 1967 census (first census carried out in the United Republic of Tanzania, when ethnic and religious identity was still being taken into account).

The Indian community in Tanganyika according to the initial enumeration exercises

Origins of immigration

  • 4 Coupland (1938, 1968).
  • 5 This figure of 6,000 is high compared to 3,000 enumerated in 1900 in Tanganyika alone. See infra.
  • 6 Prunier (1990). See Annexe III of Prunier’s publication for a historical presentation of Tanzania’s (...)
  • 7 M.F. Lofchie (1965). See the article by Marie-Aude Fouéré in this volume.

12Although the initial contacts between India and East Africa date back to at least two millennia (through sailors and traders), permanent Indian settlements are a more recent phenomenon. The largest settlements date back to the beginning of the 19th century in Zanzibar at the time of the Omani Arab sultanates. Coupland4 estimates at about one thousand the number of Indians residing on the archipelago around 1840, while Mangat5 puts them at 6,000 people in 1860 (Prunier, 19906: 14). These figures are questionable as well as controversial. Do they refer to Zanzibar (and to the neighbouring islands of Pemba and Mafia, also part of the archipelago), or to a larger area of East Africa that included the Kenyan coast? Another author estimates the number of Indians living in Zanzibar in 1859 to have been a mere 2,000 people7. All authors agree, however, that within this same region in general (from Tanga to Lamu, including Zanzibar, Pemba, Mombasa and Malindi), Indians already played an important economic role by the end of the 19th century. Few among them, nevertheless, dared to venture inland which Arab merchants almost exclusively crisscrossed at the time.

13During the German occupation (Deutsch Ost-Afrika), most Indians remained on the coast while the Germans (aware of the gains made by their British neighbours in Kenya from the manpower they provided) urged the colonial authorities in Bombay to encourage emigration to Tanganyika. This move scored little success, according to Prunier. However, the author cites the figure of 3,000 Indian pioneers in 1900, who in some cases stayed illegally Tanganyika (Prunier 1990: 238).

  • 8 United Nations, Department of Social Affairs, Population Division, Trust Territories Population Rep (...)

14Between 1900 and 1913, population growth among nationals of Indian descent was very high, reaching 9,500 in 1913 in the entire mainland Tanzania (excluding Zanzibar). This was close to double the population of non-Africans (Europeans 5,336 in the same year (United Nations 19498: 140). Some of these Indians were just contractual labourers who had no definite plans of staying in Africa. The First World War then momentarily stopped the migration.

Image 3.1 Photograph of an Indian owned shop in Tabora during German rule

Image 3.1 Photograph of an Indian owned shop in Tabora during German rule

Source: http://www.postcardman.net/​141727.jpg - Deutsch Ost-Afrika - Tabora - Indian shop.

After WWI

  • 9 It should be noted that the redrawing of borders at the initiative of the Belgian and British colon (...)
  • 10 Wikipedia, http://en.wikipedia.org/wiki/Tanzania, 15 March 2006.

15After assuming administrative control of Tanganyika in 1920 via a mandate from the League of Nations, the British encouraged new Indian immigrants to settle9. Out of the total population of Tanganyika estimated at 3,500,00010, Prunier puts the number of Indians living in the country at the time at about 10,000 while emphasizing the mobile nature of this population (Prunier 1990: 238). Ties developed between the Indians living in Tanganyika and those in Kenya and Uganda, since networking taking place within the Indian communities facilitated travel between the three territories.

  • 11 The exercises “seem to have consisted in enumerating people in certain regions, complemented by ass (...)
  • 12 “In 1921, Germans were repatriated and none of them was allowed to come back to the Territory. Euro (...)

16Knowledge about the population in Tanganyika improved gradually when the British carried out the three “censuses” we mentioned in 1921, 1928 and 193111. Due to the fact that Indians were in majority settled in urban areas – where they were neighbours with the British colonialists – and considering the role they played in the economy, they were subjected to relatively precise statistical follow-up, which was in any case easier to carry out than among the African people. In 1921, the population of Indians and Goans reached 10,209 (United Nations 1949: 140) – hardly higher than eight years earlier – but the number of Europeans had, in the interim, reduced by half due to the consequences of WWI: they were 2,447 in 1921 compared to 5,336 in 191312.

  • 13 Tanganyika Territory, 1932. Though it is not possible to access this publication in Tanzania, a cop (...)
  • 14 Thus, the Indian population is estimated to be 23,422 on pages 140 and 142 of the report, but at 32 (...)

17The 1931 census, published in 1932, indicated that 5,022,640 people were natives of Tanganyika, in addition to 8,228 Europeans and 32,398 “Asians”13. The census was silent about the existence of the Arab minority, though we know that the British had as a rule to numerate them. The 1949 United Nations report on the population in Tanganyika, itself marred by errors, suggested that the Arab population be included in the “Asians” group14. Carried out a year earlier, the 1948 census confirmed astatistical distinction between the two communities.

Table 3.1 Non-natives and African populations of Tanganyika (1948)

Populations

Total number in the territory

Percentage ‰

Part of each group (in %)

Europeans

10 648

1,4

15,2

Indians

44 248

5,9

63,1

Goans

2 006

0,3

2,9

Arabs

11 074

1,5

15,8

Others

2 184

0,3

3,1

Total Non-Africans

70 160

9,4

100

Africans

7 407 517

990,6

Grand Total

7 477 677

1 000

Source: Report on the population of Tanganyika, United Nations, September 1949.

18From this data, the size of the Indian population can be estimated for the years 1931 and 1938. Knowing the number of the European population (especially the British) and considering the Arab population to be relatively stable (because it was not so much influenced by migratory movements), the number of Indians can be established by subtracting from a total of about 21,000 people in 1931 and 22,000 in 1938.

19Table 3.2 Estimates of the population originating from the Indian sub-continent in 1931 and 1938

  • 15 Remember that our assumption is that Arabs had been assimilated to the “Asians” category in 1931. I (...)

20The figures on Table 2 should be taken with reservation15 since the Arab population is not known. They show nevertheless that there was probably little change in the size of the Indian population in Tanganyika between 1931 and 1938. In other words, the first wave of immigration encouraged by the British took place during the years before 1931, which are the initial years of the colony (1920-1931).

Structure of the Indian population

21According to the 1931 census, the sex ratio was very unequal in favour of the male population at 159 men to 100 women. The population was generally young: people aged 45 years old and above composed 9 % of the entire population while adults aged between 20 and 44 years old were 47 % and those under 20 were 44 % of the population. Women were a lot younger than men: 38 % of men were under 20 years old compared to 56 % of women in the same age bracket whose average age was not above 18 years old (25 years old was the average age for men). There was almost no stated professional activity among the women. While emphasizing relatively higher male immigration and a high number of single males, these figures indicate a possible higher mortality among young boys, as well as the fact that female immigrants were young, probably because many were young girls destined to get married.

Social and professional distribution

22The following table, prepared using the 1931 census data, describes Indian professional distribution in Tanganyika’s economy. It is evident from the table that Indians occupied an intermediate position between the Europeans and the Africans (who were not mentioned here but were at the time almost exclusively engaged in agriculture and livestock).

Table 3.3 Non-African male population between 20 and 49 years old by sectors of activity in Tanganyika – 1931 census

Table 3.3 Non-African male population between 20 and 49 years old by sectors of activity in Tanganyika – 1931 census

Source: Report on the Non-Native, Census 1931, September 1949.

23The Indians, who were a majority in business occupations like the Arabs, were also involved in industrial or administrative jobs, whereas the Arabs, less educated, were nearly excluded from these sectors. However, settled in urban areas (52 % lived in Dar es Salaam, Tanga and Tabora), they almost did not engage in agriculture, neither as farmers nor as salaried employees. European farm owners in reality employed Africans to work in the fields.

Changes in Indian population between 1938 and 1948

  • 16 There is still no distinction between Pakistanis and Indians despite the 1947 separation.

24Between 1938 and 1948, there was steady growth of the Indian population in Tanganyika, marking a break with the period of relative stagnation of the period in-between the two World Wars. Estimated at 46,254 in 1948 (including the Goans, who were 2,006), the Indian population more than doubled in a decade16. Whereas the entire population of Tanganyika in the same year was 7,477,677, the ratio of residents of Indian origin reached about 6 for every 1,000 people compared to about 4 for every 1,000 people in 1931. Indians were now 63 % of the “non-African” minorities (66 % inclusive of Goans), a percentage that did not exceed 50 % in the 1930s. With a ratio of 122 men to 100 women in 1948, the sex ratio was more balanced than in 1931. Taking into account steady migration (of young women notably), on the one hand, and fecundity among the residents, on the other, the population structure of Indians by age remained young: 50 % of the men were less than 20 years old in 1948 (38 % in 1931). Whereas the male population became slightly younger (24 years old in 1948 compared to 25 in 1931), the female population became older by about two years (20 years old in 1948 compared to 18 in 1931). In contrast, between 1931 and 1948, the European population became older by an average of 7 years (28 years old in 1931 and 35 years in 1948). The under 20 years old became fewer among the Europeans (18 % of the population) but remained more dominant among the Indians (50 %). The age of Europeans compared to Indians reflects in its own way the dominant position held by the former over the latter.

The Indian community in Tanganyika as Trust Territory under the United Nations (1947-1961) up to the first census during post-independent Tanzania (1967)

  • 17 Having attained independence in December 1963, Zanzibar united with independent Tanganyika in April (...)

25After WWII, the management of the Tanganyika territory was handed over to the United Nations, but the British held important administrative positions as well as a leading economic role until independence (1961). After the 1964 union with the Zanzibar archipelago (geographically comprising the islands of Ugunja, Pemba and Mafia), the geopolitical entity thus became Tanzania17. During this period, and the three years that followed the creation of the new Tanzanian state, three censuses made it possible to follow the demographic progression of the Indian population: in 1952, 1957 and 1967.

  • 18 “(…) The February 1952 census focused on all non-African residents as well as Africans resident in (...)

26Whereas the 1957 and 1967 censuses were termed as “general” (as were the 1931 and 1948 censuses), the 1952 census only concentrated on the urban centres (townships) and the some densely populated areas18. There is no mention of population that was not enumerated and the observer is not in a position to know whether the figures produced were “corrected” in order to reach an estimated figure of the total population. If this was not the case, then the Indian population may have been underestimated, even though the vast majority of this population resided in urban areas. The released figures, nevertheless, represent intermediary figures which are consistent with the earlier census (1948) and the census that followed (1957). Cf. Tables 5 and 6.

Some data on migration

  • 19 The high influx of Europeans (mostly British) entering and leaving Tanganyika may also be noted. Ho (...)

27In the first place information is scarce to assess the significance of the migration influx between the Indian sub-continent and Tanganyika, though it is sufficiently precise to give way to interpretations. The figure of 46,254 has been previously identified as the population of Indians and Goans enumerated in 1948 (44,248 + 2,006). The same year, 1,723 “authorized” or “legal” immigrants from the Indian sub-continent arrived at the country’s border posts, thus bringing the annual rate of immigration to the high percentage of 37 %. Using the data on the influx from the three previous years (1949, 1950, 1951), we observe however that the number of Indian residents leaving the country at that time was also as high, even higher than the number of those entering the country. Table 3 thus shows a negative rate of net migration for 1949 and 1951. Several explanations have been advanced to account for this turnover. The first has already been suggested in the 1949 United Nations report: some immigrants, having planned from the onset a temporary stay in Africa, went back to India to spend their retirement there (United Nations, 1949: 88). Another explanation is that some of the Indians left Tanganyika to go to other places in East Africa (or Southern Africa), some of them “trying their luck” in several British colonies in the region. The assumption of dynamic migration that corresponded to migrants’ mobility between countries can be verified by examining border police statistics, or by using the censuses which give indication of the immigrants’country of origin. In the case of former Tanganyika, the only information available is the 1951 data. That year, only 58 % of the 1,932 Indians who entered the country came directly from India or Pakistan. The rest of the migrants (seven out of ten holding a British passport at the time) indeed came from other African territories: 23 % had previously stayed in Kenya, 6 % in Uganda and 12 % came from Zanzibar, a territory which was still independent at the time (the remaining 1 % coming from other places). Thus at the beginning of the 1950s, Indians significantly travelled between the three main territories of East Africa. The figures show that once the attachment to the Indian continent had been broken, none of the East African countries was a priority option for immigration, except in a few exceptional cases19.

Table 3.4 Entries and Exits in Tanganyika between 1938 and 1957

Table 3.4 Entries and Exits in Tanganyika between 1938 and 1957

Source: East African Statistical Department (from records of Immigration and Passport office). Cf. from 1938 to 1948, Goans are included in the Indian population.

28Secondly, as is characteristic of many migration influxes, the trend of the unequal sex ratio that was unique to the Indian immigration was unexpectedly reversed between 1950 and 1957. Whereas from 1950 to 1955, the male population remained over-represented, an inverse trend began from 1955 onwards. Such change could suggest that matrimonial strategies were deployed by some of the men who had come alone to Tanganyika and may have invited their wives or future wives to join them once they were economically stable.

29The age structure of the new residents in 1957 makes this assumption probable. Besides, more women than men immigrated to Tanganyika in 1957, 482 of them aged between 15 and 29 years old while the men in the same age group were 280. The data corresponds to a sex ratio of 58 (58 men for every 100 women). As for the age group between 30 and 59 years old, the sex ratio is as high as 112.

Evolution in number and age structures

30Although women who settled in Tanganyika after 1955 were more in number than men, as just seen, the entire Indian population remained dominantly male in every census. In 1952, 1957 and 1967, the sex ratio was 118, 111 and 110 respectively. In other words, the intensity of female migration was not enough to rebalance the earlier numerical advantage of the pioneer male population.

31At the same time, the steady growth of the Indian population in Tanganyika between 1931 and 1948 continued – and even heightened – up to 1957 when it reached a population peak of 76,517 (Table 5). Between 1957 and 1967, the cumulative numbers from births and migration could no longer compensate for the numbers from deaths and departures. The population size therefore decreased but at a low average percentage of -0.2 % per year.

Table 3.5 Demographic growth of the population of Indian origin in Tanganyika / Tanzania between 1931 and 1967

Years

Population of Indian origin in
Tanganyika (or Mainland)

Average growth rate per year

1931

(20 898)

1948

46 254

4,8 %

1952

59 739

6,6 % (*)

1957

76 536

5,1 %

1967

75 015

- 0,2 %

Cf. The average growth rate per year of the population of Indian origin living in Tanganyika is 4,8 % between 1931 and 1948.
(*) Seeing that the 1952 census did not cover the whole territory, it is impossible to know if the values estimated are estimated for the whole of Tanganyika or if they under-evaluate the Indian population. In that case, the average annual growth between 1948 and 1952, which is the highest during the 1931-1967 period, would even be higher (estimated at 6,6 %).

32In addition, the percentage of the Indian population in Tanganyika was highest in 1957: Indians in the large sense of the word (including Pakistanis and Goans) were then 0.87 % of the colony’s population. This percentage fell to 0.73 % ten years later. However, compared to the foreign population, and because of the high population growth among the Europeans and Arabs as indicated by the census, the percentage of Indians decreased between 1948 and 1957.

Table 3.6 Evolution of the population of Indian origin between 1931 and 1967

Table 3.6 Evolution of the population of Indian origin between 1931 and 1967

Sources: Censuses of Tanganyika and Tanzania

  • 20 Even though, from a geographic point of view, it was part of the Zanzibar archipelago, Mafia Island (...)

33When Tanzania came into being in 1964, 13,552 “Asians” enumerated in Zanzibar added to the Indian population in Tanganyika (the territory that became referred to as the “Mainland” in the United Republic of Tanzania). The inclusion of the figures caused a somewhat artificial increase in the general figures. For Zanzibar as a whole without any other retrospect (although there was census in the archipelago in 1948 and 1958), it is impossible to determine whether there was also a decline in the Indian population in the Unguja and Pemba islands after 195720. What is sure is that the 1967 census data shows that the Indian presence in Zanzibar was then distinctly higher in proportion to the population than on the Mainland (about 4 % instead of 0.7 %, a six-fold higher). This difference can be explained by the migration to the two islands that dates back to earlier years than on the Mainland (former Zanzibari trade posts). This is what comes out of the comparison of Indian age structures in Zanzibar and on Mainland. Indeed, the ratio of people aged 60 years old and above is 4 % on the mainland while it is double this percentage in Zanzibar: the population was older there because it had had the “time to age”, having arrived much earlier.

34In the 1950s and 1960s, the Indians living in Tanganyika were a relatively young population, even though it no longer had the profile of a pioneer population. In 1957, the average age of the entire population was 22.2 years old, with some differences between the three communities. Indians and Pakistanis had very similar age structures: the average age was 22 years old for the Indians and 21.5 years old for the Pakistanis. As for the Goans, they were generally older owing to a lower ratio of the under 20 year-olds (average age being 24.9 years old).

35In the three communities, men were older than women: +1.5 years old among the Pakistanis; +2.1 years old among the Indians. The gap was a lot more significant among the Goans: on average, men were 5.3 years older than women.

36In 1957, the “Asian” age structure on the whole (Indians, Pakistanis and Goans) was hardly different from that of 1952. Even though the 1952 data is insufficient to calculate an average age, it can be observed that the ratio of the under 20 year-old youth rose to 53 % for the three communities, a figure similar to the one of 1957.

37The Indian age structure actually began to alter after 1957 due to the decline in the migration influx and probable change in fertility behaviour. Within a space of ten years, ageing began within the population: the “Asian” average age in Tanganyika ( “Mainland”) reached 32 years old: the average age in the 1967 census was 25.4 years old on the Mainland and 26.6 years old in Zanzibar. These demographic changes are testimony of the Indian population’s uniqueness in comparison with the African native population and the long-term settlement of a higher number of immigrants.

38The above diagram reveals the “pyramid” nature of the age structures in Tanganyika and Tanzania, starting with the African population in 1967. This reflects a pre-transitional demographic clustering, with the population born in Tanganyika recording high birth and mortality rates (hence the typical pyramid shape of the African populations at the time, observable even today in some countries on the continent). In 1967 (like 1957), the Indian age structure profile is different from the natives’: the ratio of the under 10 year-old youngsters is lower, while that of the 10-59 year-olds is much higher. These variations emanate from three main demographic phenomena: mortality, fertility and migration. The “migration” roots make Indians a “select” population. At any given time and in any given area, it is known that young adults are the section of the population with the highest propensity to emigrate. In 1957, for example, immigration among over 59 year-old “Asians” was low (3 % of the cases). In contrast, the ages of close to half of the new immigrants (47 %) were between 15 and 34 years. These numbers do not take into account the fact that a section of the young adults could have come along with their young children. In addition, mortality and fertility also distinguished “Asians” from Africans. Enjoying an intermediate status between the British and the natives, the Indians were more educated than the natives, had a higher life expectancy, and endeavoured much more to control their lineage. Between 1957 and 1967, the base of the age pyramid narrowed while the ratio of Indian adults increased. This was a sign of indisputable decline in fertility (Ref. Figure 3.1).

Religious belonging among Indian communities in Tanganyika

39The 1957 census provides information on the religion of the Indian people enumerated. Most of them were Muslims (46.0 %) or Hindus (44.4 %) while the Sikhs came third (6.5 %). Those who indicated they had come from Pakistan (having migrated before the 1947 separation) were nearly all Muslims. As for the Goans, who had been under the Portuguese cultural influence, they were almost exclusively Christian (Catholics).

  • 21 This table does not fully reflect the cultural and religious diversity of Indian communities. Hindu (...)

Table 3.7 Distribution of immigrants of Indian origin in relation to their religion in the 1957 census21

Religion

Indians

Pakistanis

Goans

Christian

886

15

4 732

Hindu

29 035

3

10

Islam

30 082

6 272

9

Jain

913

0

0

Sikh

4 232

3

0

Parsee

170

0

0

Jewish

5

0

0

Other

11

0

0

Unknown

31

2

6

Total

65 365

6 295

4 757

Geographical distribution of people of Indian origin

40In 1967, there were 88,567 people of Indian origin in newly created Tanzania (15 % living in the archipelago of Zanzibar and 85 % on the Mainland). This population mainly lived in urban areas: on the mainland, 78 % of all those enumerated resided in an urban-classified area as compared to 4.3 % of the African population, 25 % of the Arab population and (only) 34 % of the European population (there is no urban/rural distinction for Zanzibar). The Asian population’s urban tropism can be explained by the specificity of the immigrants’ professional occupations in sectors linked to urban activity: cottage industry and manufacture; administrative jobs, business, services… The same attraction to the cities can be observed among Indians in Kenya and Uganda.

  • 22 These are ratios calculated using the 1957 (and 1967) census data.

41With 36 % of the Indians in the country concentrated in Dar es Salaam alone, the ratio of this population in the former capital in 1967 was equal to 11 % of the total population. The different communities were, however, represented differently in the capital: one Goan out of two lived there and only one Pakistani out of four22. Due to high economic and business potential, proximity to borders and infrastructure linking them with the capital, other urban centres were also destinations of choice for the Indian communities. This was the case along the country’s northern coast in Tanga (7,000 Indians to 1,000 Africans in the Tanga region). Near the Kenyan border and close to Nairobi, the cities of Moshi and Arusha, which were important tourist and agricultural centres (coffee), constituted meeting points for Indians from the two countries. Mwanza, situated south of Lake Victoria, received many Indian businessmen attracted by the region’s wealth (cotton, coffee) and its proximity to several borders (Uganda, Kenya, Rwanda, Burundi). Albeit in smaller minorities, the then rural townships of Tabora, Dodoma and Morogoro in central Tanzania, which were connected to Dar es Salaam by the railway, had Indian traders during colonial times, whether retailers or wholesalers. Though heavily dominated by Arabs (25 % of the population), Indians also had a strong presence in Pemba with a ratio of up to 1.4 % of the total island population (and 2 % of all Indians in Tanzania).

Where is the Tanzanian Indian population today?

  • 23 High Commission of India (http://www.hcinairobi.co.ke/). Indian diplomatic officials generally say (...)

42The counting of Indians in Tanzania thanks to public statistics stopped in 1967. Forty years later, it has consequently become impossible to measure the size of this population through national census – whether some of them had become Tanzanian citizens ( “Indian Tanzanians”) or, for more recent migrants, kept their foreign nationality (Indian, Pakistani, British, Kenyan, etc.). In 2006, the diplomatic services of the Indian Embassy in Nairobi23 estimated at about 45,000 the number of Indians living on the Tanzanian soil. This is half the figure of 1967. A large majority of them (40,000) had Tanzanian nationality, having mainly come from Gujarat and, as in the past, residing in the country’s cities and urban centres (Dar es Salaam, Arusha, Dodoma, Morogoro, Zanzibar, Mwanza, Mbeya) – but to a lesser extent in Zanzibar, which many Indians, particularly Ismailis, neglected. The remaining 5,000 held Indian nationality (the other nationalities previously mentioned in 1967 are not referred to).

43In order to establish the number and composition of the Indian population in Tanzania today, other forms of investigating would have to be conducted, such as deeper investigation among the representatives of religious and status communities, enumeration of pupils in schools, statistics on residence certificates). While assuming that such research methods can be conducted at the national level, it should be remembered, however, that they will never guarantee the level of accuracy that a national census can (potentially) attain.

Indian diaspora in Kenya

  • 24 In contrast, all published statistics are freely accessible. Various censuses conducted by the Brit (...)

44Indian presence in Kenya beyond the coastal region, which remained insignificant until the end of the 19th century, strengthened following the construction of the railway from Mombasa to the source of the Nile in Uganda. Contrary to the situation observed in Tanzania where public statistics authorities have for several decades stopped numerating Indians, information based on the criterion of origin is, as mentioned above, still collected in Kenya but it is no longer released. It is therefore still possible to access detailed knowledge on demographic changes, subject to permission to access all the sources24.

Kenya’s Indian population in retrospect25

  • 25 For a more complete history in French, see Prunier (1990).

45The first great wave of Indian migration to Kenya was engineered by the British at the turn of the 19th century during the construction of the Mombasa-Lake Victoria railway. However, out of the 32,000 contract workers recruited in India during this period, only 2,200 remained in the country, while about 2,500 died during construction. Considering the very low number of Indians already in the country by then – on which there is, however, no digital data – the number of Indian settlers in Kenya therefore swelled before the completion of the railway by other migratory influxes whose specific chronology and nature remain unknown.

Image 3.2 Photograph of the Mombasa–Lake Victoria railway under construction

Image 3.2 Photograph of the Mombasa–Lake Victoria railway under construction

Source: Museum of Nairobi, temporary exhibition on Indian heritage (January 1995) Some Indians in the middle of the photograph supervise the work of Africans.

46The first British enumeration carried out in 1911 came up with a figure of 11,787 Indians in Kenya compared to 3,175 Europeans. The Indian population in Kenya doubled between 1911 and 1921 and also during the decades that followed to attain 43,623 people in 1931, 97,687 at the end of World War II (1948 census), and 176,613 in 1962, on the eve of independence. Indians, who mostly held British nationality, were then three times more numerous than Europeans.

Kenya’s Independence

47After an African government came to power in Kenya, Indians experienced discriminatory measures aimed at limiting their business activities, and generally weakening their socio-economic influence (such as the Trade Licensing Act). At the end of 1967, many people began to migrate from the country (mainly to Great Britain) as a premonition of the events that occurred five years later in Uganda following the expulsions and systematic expropriation. In the 1969 census, the Indian population in Kenya had decreased by 37,500 people. The exodus continued during the following years, with thousands of people leaving their adopted homeland without plans of coming back. The number of Indian settlers was halved within fifteen years, attaining a total 78,600 individuals only in 1979.

48However, under Daniel arap Moi’s presidency, the Indians who had remained in Kenya found themselves in a more satisfactory social and economic standing and a reverse migration trend began, either due to the return of former settlers or fresh arrivals in the country. Spread throughout in multiple areas of activity, Indian investment played an important role in modernizing the country. Even though time was ripe for renewed demographic growth of Indian communities, it remained very moderate. Thus the population of Indian settlers in Kenya attained 89,000 in 1999, a figure just above the number recorded thirty-seven years earlier in 1962.

Figure 3.2 Evolution of the Indian population or the population of Indian origin living in Kenya (1840–1999)

Figure 3.2 Evolution of the Indian population or the population of Indian origin living in Kenya (1840–1999)

Sources: CBS, Kenyan population census

49During the entire 20th century, population growth among the Indians followed by a decline after November 1967 changed their place in the national population or in comparison to the two other important non-African communities in Kenya: Europeans and Arabs. There is substantial data attesting such changes seeing that the British, as previously mentioned, always conducted census of the so-called “non-native” populations of Kenya separately ( “Non-Native Population of Kenya Colony and Protectorate”). The separate count rule continued after independence, with the new census authorities calling it the “non-African” population category.

  • 26 Restricting the issuance of the lion’s share of business licences to Kenyan citizens.

50From 1911, Indian population (together with Goans, who were considered separate from Indians until the 1949 census) was the highest of the three “non-African” groups, accounting for 50 %. This proportion attained 2/3 of non-Africans between 1948 and 1969, in spite of the wave of Indian departures following independence, but was compensated by departures of European settlers as well as Arabs, who were also affected by discriminatory provisions of the Trade Licensing Act26. During the entire 20th century, the Indian population remained two to three times higher than the Europeans.

51Being the first minority group of the country, Indians were and are still noticeable in the population because they mainly live in urban areas and are concentrated in certain districts. However, during the 20th century and despite the growth in their numbers until 1967, their demographic weight in Kenya was eroded due to high growth rate of the African population. There were 20 Indians for every 1,000 Kenyan inhabitants in 1962, less than 13 in 1969 and only 3 in 1999 (and this ratio is still decreasing today).

Table 3.8 Evolution of the three main non-African populations living in Kenya between 1911 and 1989

Table 3.8 Evolution of the three main non-African populations living in Kenya between 1911 and 1989

Sources: The East African Statistical Department until 1948 then the Statistics Division.

52As for regional distribution, the two largest Kenyan cities, Nairobi and Mombasa, were home to the majority of the country’s Indians in 1948 (about 70 %). Whereas during the entire first half of the 20th century, the number of Indians in Mombasa was close to the number in Nairobi, there was a significant decline in their numbers in Mombasa in the second half of the century, with close to half of the Indians residing in the capital compared to only a quarter in Mombasa. These ratios were still evident at the very end of the century (1999), as will be observed in the following section.

Contemporary Indian population in Kenya

  • 27 It has not been possible to update this article by taking into consideration the last census.

53In view of the above retrospect, we can now zoom in on the characteristics of the “Indian” population currently living in Kenya (2008), using data of the 1999 census in comparison with those of 1989 (the next Kenyan census is planned for 2009 and detailed data collected will not be available before 201127).

54The 1999 census made it possible to account for close to 90,000 Indians living in the country (Ref. Annex 1). This figure included both holders of Kenyan citizenship, holders of foreign passports (Indian, Pakistani, United Kingdom, etc.), and people classified “other Asians” of unspecified citizenship, but probably originating from former Indian possessions of the British Empire (Sri Lanka, Bangladesh, etc.). These statistics obviously cannot be taken as reflecting any exhaustive enumeration. Even without necessarily questioning the quality of data collection in Kenya, the capacity to gather information on all individuals is hindered by objective conditions defining the specific situation which affect some of them (geographical isolation, illiteracy, etc.). In addition, as in every country of the world, people who want to conceal their presence and who believe they can be identified through a census try not to answer such questions. This is the case of those who fear sanctions of any nature from the regulatory authorities (such as people concerned with escaping the tax regime or previous fiscal fraud, contravention of labour laws, illegal entry into the country, overstaying the authorized visa period, etc.).

  • 28 The remaining 40 % are “Asians” who had moved from one district to another within Kenya.

55Even though, for the above-mentioned reasons, it is highly plausible that part of the population in Kenya was not enumerated (which particularly comprises immigrants who just recently entered the country), it remains difficult to assess its scope. It should be noted, however, that according to the 1999 census, the number of enumerated “Asians” who were not Kenyan residents one year earlier (1998) was as high as 7,036, close to 8 % of the total population. This mobility, which can be attributed to external factors, is therefore relatively high. Moreover, it accounts for 60 % of the Indian residential mobility between 1998 and 199928.

  • 29 A figure between 3 and 8 % seems more realistic.
  • 30 Supposing, in a totally arbitrary manner, that one entry out of four is illegal, leading within fiv (...)
  • 31 Particularly Michel Adam, who advances a figure of 120,000 for the year 2002 based on approximate e (...)
  • 32 In its publication Population and Housing Census, Analytical Report on population Dynamics, vol. II (...)

56As high as these figures may be, it appears nevertheless unlikely that the number of unaccounted “Indians” in Kenya is higher than 10 % of the community’s total population29. The conclusion of the latest census is that the entire Indian population in 1999 did not exceed 100,00030, a figure that, it should not be forgotten, should be compared to the highest figure attained thirty-seven years earlier in 1962 (about 200,000 people taking into consideration an equivalent percentage of unaccounted for individuals). Even though this figure appears definitely lower than some authors’ estimates31, the fact remains that “Asians” are still today the largest non-African minority in the country, with about half of this population being Kenyan nationals32.

Three quarters of Indians still concentrated in Nairobi and Mombasa in 1999

  • 33 Locations in Kenya are administrative units that are equivalent to the French “communes”. Nairobi i (...)

57The 89,310 “Asians” enumerated in 1999 in Kenya are very unevenly distributed all over the country. Just as it has always been in Kenya, and everywhere in East Africa, this population essentially lives in urban areas (94.9 % urban areas; 1.3 % non-urban areas), with the percentage of those resident in rural areas hardly above 3.9 %. Half of the population (50.5 % or 45,088) are concentrated in the Nairobi area alone while a quarter (26.8 % or 23,894) live in Mombasa. The remaining quarter is spread in 67 other districts around the country. Apart from Nairobi and Mombasa, four districts have between 1,000 and 3,340 Indian residents: Kisumu, Nakuru, Uasin Gishu (Eldoret) and Thika, while in 35 districts (one out of two) resident members of the “Indian” community are less than 100 people (26 out of 50). It is worth noting, however, that out of the 2,432 locations in Kenya, 1,084 of them – close to half – host members of the Indian diaspora (see maps in Annex 3)33. Even though the Indian population is highly concentrated in some parts of the country and despite being small in absolute numbers, few Africans are completely unaware of their presence.

58The maps below summarize graphically show the concentration of Indians in some districts, especially the specificity of their distribution throughout the country vis-à-vis the distribution of the entire Kenyan population.

Image 3.3 Maps of Kenyan districts, total population and Asian population per district

Image 3.3 Maps of Kenyan districts, total population and Asian population per district

Source: UBOS – Kenyan 1999 Census Cartogrammes were made with the help of Dominique Andrieu (Tours, France)

59Map 4.1 shows the administrative outline of Kenyan districts. The unequal surface areas covered by the districts, particularly the very smaller ones in the west (around Lake Victoria), and the larger districts in the north, towards Sudan and Ethiopia, is immediately noticeable. Generally, this spatial inequality is inversely proportional to population density: a district with high density cannot be administered properly if its area is small. In addition, there is less interest in producing chromatic maps based on the administrative borders, because the eye tends to be less attracted to least populated areas. The cartogram technique which artificially distorts the contours and the surfaces of the districts (maps 4.2 and 4.3) makes it possible to go round this difficulty and to better identify the districts with higher populations. Thus, the 4.2 view shows what the map of Kenya would be if the areas of the districts were proportional to their population size. The “small” districts (geographically significant) in western Kenya are therefore very easy to find, just like the capital, Nairobi, which was nearly invisible on the first map.

60On the same note, the third view ( “Indian” population cartogram) indicates the very unique nature of this community vis-à-vis the entire Kenyan population. The two centres, Nairobi and Mombasa, appear to be ultra-dominant as other districts fade out. Only Kisumu and Eldoret, to the west, provide noticeable secondary centres. This diagram confirms the very urban nature of the “Indian” population distribution.

61In the 1999 census, even in the districts that have the highest number of “Indians”, the relative weight of this community is light. In 64 out of the 67 districts, they account for less than 0.5 % of the total population. Though fewer Indians live in Mombasa than in Nairobi, the community’s relative weight is heavier in the coastal city, accounting for 3.6 % (1 in every 28 residents). In the Nairobi areas, the percentage “Indians” is hardly above 2 % of the total population (2.1 % or 1 out of every 48 residents).

  • 34 This actually refers to Nairobi West estate, which is part of Kibera Division. For Nairobi resident (...)

62Inequality in the Indian population distribution between districts is accentuated by inequality in distribution within the districts. In Nairobi, the divisions (sub-units of districts) in which the largest number of Indians are concentrated: Westlands (50.2 % of all “Asians”), Central Nairobi (28.6 %) and Kibera (13.5 %)34. In Mombasa, “Indian” communities are concentrated in the central part of the city (84.4% in the “Island” division); Kisauni district comes a distant second with 13.2 % of Indians.

63This type of geographical distribution shows that the people concerned opt to settle in communities in relation to their economic activities or their social status. Bernard Calas argues that these spatial settlements make up “racial and social” territories insofar as Nairobi is a “socially compartimentalized” city (Calas 1998: 31, 47). Some authors also emphasized the low level of social relations between “Indian” communities and the African population, a situation that does not encourage social and ethnic cohabitation in the residential estates. In Nairobi, as everywhere in the world, the average price of real estate property within the locations and divisions is also likely to give the researcher a clear idea of the social roots of its residents.

Image 3.4 Map of Nairobi district – Indian Population by « location » (1999)

Image 3.4 Map of Nairobi district – Indian Population by « location » (1999)

Source: CBS - Kenya 1999 Census

64The map shows Nairobi District (surrounded by Kiambu District to the west, Thika to the north-east, Kajiado to the south and Machakos to the east) and the various locations that are part of it. The pattern underlines Indian presence in every location and outlines eight in which more than 2,000 Indians were enumerated in 1999. By descending order, they are Highridge, Ngara, Nairobi West, Parklands, Kariokor, Kilimani, Lavington and Starehe. These eight locations are home to 89 % of the Indian population or of Indian origin resident in Nairobi area.

Table 3.9 « Locations » of Nairobi district having more than 2000 Asian inhabitants –Socio-economic indicators.

Table 3.9 « Locations » of Nairobi district having more than 2000 Asian inhabitants –Socio-economic indicators.

Source: CBS - Kenya 1999 Census

65Some socio-economic indicators make it possible to distinguish the “locations”.

  • In Kibera Division, the profile of residents in Nairobi West location is close to that of Indian residents in the entire Nairobi District.

  • In the four locations of Westlands in which the largest part of the “Indian” population is concentrated, the sex ratio is more or less balanced and a big fraction of the residents belongs to higher social categories. The ratio of those who went to university is distinctly higher than elsewhere, particularly (minority population) in Kilimani and Lavington locations, which were residential areas for most Europeans. Among the eight locations in Table 9, the ratio of held assets is highest and the number of jobseekers is the least in Kilimani and Lavington. For their part, Highridge residents, who constitute half of all “Indian” residents in Westlands Division, seem be settled there for a long time because most of them declare they are Indian Kenyans; close to half of them were born in the district and retired people are more numerous than elsewhere.

  • The Downtown (Central) Division hosts Indians belonging to lower social categories. A comparison of the three residential areas reveals other common points with residents of Starehe and Ngara areas, which have a high proportion of recent immigrants (highly unbalanced sex ratio, high percentage of people born in India). Kariokor residents, with an equivalent social status (even much lower than that of Ngara residents seeing the low percentage of those who have gone to university), have different characteristics. Settled for a long time in Kenya (a sex ratio of nearly 100), Kariokor residents are mostly Indian Kenyans (with the lowest proportion of people born in India). Economically, they appear more vulnerable, with less held assets and a high percentage of jobless people.

Places of birth of Indians living in Kenya

66Two thirds of people of Indian descent enumerated in 1999 were born in Kenya (there is caution with regard to the number of those born on the African continent). It can be assumed that having kept major cultural traits of their ancestral land, they changed some of its aspects due to the spatial distance and initial socialization that did not take place neither in India nor in Pakistan. Despite the small size of their community – compared to the African population – these people are now part of the East African human reality.

67Curiously, the percentage of “Indians” born in Kenya may have nevertheless decreased between 1989 and 1999 from 75 % to 66 % of the total “Indian” population, whilst the total figures between the two censuses seem to have been somehow stable. Even though this data can be interpreted as a sign of increasing mobility within the diaspora, it should not be ruled out that it could have been released as a result of typing errors, as several errors identified in the 1999 census may lead to suspect (see Annex 1: confusion between Sudan and India and lack of data on natives of non-African countries in 1999).

Figure 3.3 Birth place of people of Indian origin living in Kenya in 1999

Figure 3.3 Birth place of people of Indian origin living in Kenya in 1999

Indian populations in urban centres

  • 35 Remember that we identified the Indian population living in Kenya using the P13-Tribe census variab (...)

68Although concentrated in some urban centres, “Indian” populations in major Kenyan cities do not show the same characteristics, which are related to the history of Indian migration in East Africa and with the stated centrality of the capital. The “Indian” population in the coastal city of Mombasa is the most unique, with 71 % of people of Indian descent in 1999 declaring they were Kenyan-Asians. In Nairobi, this percentage did not exceed 44 % while in Kisumu, a city on the shores of Lake Victoria, the percentage was only 33%35. Conversely, the percentage of people who said they were “Indian” in Mombasa (22 %) is half of those who said the same in Nairobi (46 %) and nearly a third of those in Eldoret (56 %) and Kisumu (59 %).

69These differences cannot be understood without reference to the initial years of the diaspora. The coastal region of Mombasa (including the former trade posts of Malindi and Lamu) was the first to receive immigrants. There is little surprise therefore that the area has the lowest percentage of “pioneer immigrants”. Moreover, Mombasa is no longer the “gateway” to the country since most of the travelling is by air. Access to the country is now through Nairobi, which plays the role of “turntable” for the migratory routes. It is after a rather long stay in the capital that the new newcomers make the eventual decision to plunge into further migration towards smaller towns, such as Kisumu, Nakuru or Eldoret. By matching places of birth with places of residence in 1999, one can assume the degree of autochthony of individuals. Close to three out of four Indians living in Mombasa are born in the city, whereas this proportion falls to less than one out of two in Nairobi and less than one out of three in Nakuru and Eldoret.

Table 3.10 Birth place of Asians Living in Nairobi, Mombasa, Eldoret, Kisumu and Nakuru.

Table 3.10 Birth place of Asians Living in Nairobi, Mombasa, Eldoret, Kisumu and Nakuru.

Source: CBS - Kenya 1999 Census

70In Mombasa, Indians born in India are also significantly less numerous in proportion (10.8 %). They are three times more in Nairobi and four times more in Eldoret. Among the five large urban centres that have a high concentration of members of the Indian diaspora, Eldoret appears to have received the most recent immigrants.

Socio-demographic characteristics of Indian population are distinct from African population

Fertility rates

  • 36 Source: INED (1999). It should be noted that in the “Demographic and Health Survey”, the cyclical f (...)

71According to the 1999 census, Indian women and girls of twelve years old and above said they had given birth to 0.81 boys and 0.77 girls (a “consistent” masculinity ratio of 1.05). These figures, which cannot constitute reliable birth rate, and less fertility indicators, nevertheless demonstrate a lower birth rate compared to the general level of fertility in Kenya. In 1999, our estimated gross birth rate was less than 14‰ for “Indians”, and around 38‰ for Kenya. Even though fertility tended to decrease in the country, it stabilized around the same time at around 5 children per woman36. Fertility among women from the Indian sub-continent was therefore lower than those of Kenyan women (on the whole); it also proves to be lower than the rate observed in their ancestral land (in 1999, India had a cyclical fertility indicator equal to 3.4 children per woman, while Pakistan had 5.6).

Household Structures

  • 37 “Heads” meaning “heads of family”.

72The Indian diaspora had 21,432 households in 1999, 87 % of them of male heads37. By dividing the number of enumerated individuals within the same category ( “Asians”) by the number of households, it is possible to deduce the average size of households, which stabilized at 4.2 people in 1999. Compared to African populations, “nuclear” type households are more widespread. However, such households are far from being exclusive since close to a quarter of enumerated people within the Indian community (24.1 %) live in households without being neither family head nor household wife or children. They can be relatives, brothers and sisters of the head of the household or his wife, cousins, (…) or of non-family “acquaintances”.

Table 3.11 Relationship with the referent person (household head) according to gender

Table 3.11 Relationship with the referent person (household head) according to gender

Source: CBS – Kenya 1999 Census

Remarkable age structure

73Owing to the history of migration and a relatively low fertility, the age structure of the Indian population or the population of Indian origin is quite different from that of the Kenyan population. As shown in the diagrams below, the age structure of the Indian diaspora resembles that of a Western country with a base percentage smaller than the median percentage.

Figure 3.4 Age structure of the population of Indian origin in Kenya in 1999

Figure 3.4 Age structure of the population of Indian origin in Kenya in 1999

Figure 3.5 Age pyramid of the Kenyan population in 1999

Figure 3.5 Age pyramid of the Kenyan population in 1999

74In the Indian population or the population of Indian origin, those less than 25 years old are slightly more numerous than those aged between 25 and 60 years old, a trait that does not only result from the decrease in fertility but also from demographic peculiarities in the immigrant population. Current migratory influxes, which are even less important than in the past, add to the adult age pyramid (young adults more often than not). Beyond 55 years old, there is a rapid decrease in figures because of higher mortality among the elderly (higher here than in Western countries), and also because immigrants are rarely elderly people. Contrary to this distribution, the age structure of the Kenyan population (including “Indians”) is a “developing country” type age structure, which has a much larger base (high birth rate) and a rapidly narrowing pyramid towards the summit (high mortality): those less than 25 years old are two thirds (66 %) of the total population.

Level of education is much higher among “Asians” than in the Kenyan population

  • 38 For the rates in Kenya, we calculate the figures using data from: KENYA 1999, Population and Housin (...)

75Education is another major difference between the Indian population or population of Indian origin and the entire Kenyan population. The first comparison, between “Asians” and “Kenyans” in general (including “Asians”), brings out a sharp contrast between the two categories. Only 6.2 % of “Indians” have never been to school compared to 18.2 % of Kenyans; 15.3 % of “Indians” have gone to university compared to 0.8 % of the country’s entire population38. However, the sociological relevance of such a comparison is limited because of the disparity in access to school in rural areas (where 70 % of Kenyans live) and urban areas (where 95 % “Indians” live). In this respect, comparison between the schooling rate among Indians and the schooling rate among Kenyans resident in urban areas shows – as we will see a little further ahead – results with less contrast (this would be more so the case if the comparison involved urban African families at the same social level).

  • 39 It is worth noting with some surprise that, among the Asians, more boys and girls between 6 to 9 ye (...)

76As shown in the tables below, the schooling rate among “Indian” children compared to Kenyan children is higher from primary school and the gap is especially wide between 15 and 24 years old. For the ages that correspond to university education, about two times more “Indian” boys receive education within the reference urban population (2.5 times for girls). There are differences in schooling between boys and girls within the “Indian” population (to the detriment of girls), but they are less significant than within the Kenyan population, including all ethnic groups. In the category of people who have never been to school, the percentage of “Indians” is still the lowest, for men and women alike39.

Table 3.12a Proportion of the Indian population in Kenya according to education status in 1999

Table 3.12a Proportion of the Indian population in Kenya according to education status in 1999

Table 3.12b Proportion of the urban Kenyan population according to education status in 1999

Table 3.12b Proportion of the urban Kenyan population according to education status in 1999

Source: CBS - Kenya 1999 Census

77People of Indian origin have higher schooling rates or have gone to school more than Kenyans residing in urban centres. Consequently, their literacy levels are also on the average higher than in the reference population. Even in Nairobi, the capital city and city of most universities, the residents’ (as well as “Indians”) literacy levels are lower than that of the entire “Indian” population. Among residents who have completed their initial training, 4 % have attained university level (completed undergraduate and above) while 2.3 % have attained secondary education at high school level (Forms 5-6). These percentages are 13.2 % and 5.4 % respectively within the Indian people or people of Indian origin.

What are the latest changes in the Indian population in Kenya?

78According to the digital files we accessed from the Central Bureau of Statistics (CBS) – whose quality is appreciated in Annex 1 – the Indian population increased slightly between 1989 and 1999 (an estimated 86,205 individuals in 1989 and 89,310 in 1999, representing a growth of 3.6 % in 10 years). However, official publications from the 1989 census results are not entirely consistent with these files, since the literature shows a figure of 89,185 people of Indian origin. If the latter figure were to be retained, we would speak of virtually stable figures instead of moderate growth during the inter-census period. The issue of total figure therefore remains debatable, but the gaps between the sources are relatively narrow. Moreover, whatever statistics are retained, using the files of the last two censuses enables us to give an insight of the dynamic aspects of recent demographic changes in the Indian populations in Kenya.

Positive migratory balance among young adults despite demographic ageing

79We may begin by considering changes in the age structures. The diagram below highlights changes in the weight of every age group between 1989 and 1999 for men and women.

Figure 3.6 Variation in the weight of age group between 1989 and 1999

Figure 3.6 Variation in the weight of age group between 1989 and 1999

80A distortion is observable in the age structure that generally contributes to demographic ageing within the population: increase in the weight of every age group from 50 years old for men, and 40 years old for women. In contrast with this ageing, the rise in number within the 20- 29 age group results in a higher percentage of young adults in the total population. As long as ageing by ten years between 1989 and 1999 of the young 10-19 age group alone cannot explain its growth in the entire population, the situation that the diagram shows will be interpreted as the result of migration, whose net balance favours Kenya. On the contrary, despite the presence of young adults at family starting age, the representativeness of the figures from 0 to 19 years old decreases, which can be explained mechanically by the distortion of the structure of advanced ages, as well as the relatively low fertility among people of procreating age. The growth in weight of women between 40 to 50 years old is also observable. This growth cannot be explained by ageing among the women alone.

Increase in number of “Indians” at the expense of “Asian Kenyans”

  • 40 During the census, the question asked was: “What is this person tribe?” It was recommended to the c (...)

81Another interesting change between 1989 and 1999 was the change in the structure from the point of view of the “tribe/nationality” variable. During censuses, this variable presents an opportunity for indigenous Africans to state their ethnicity40. It can also be a means for people of Indian origin to indicate their Kenyan nationality (identity expressed by the term “Asian-Kenyan”) although – since it is not an obligation – this declaration cannot be considered a matter of fact (this issue has been discussed in Annex 1). People of Indian origin can also say they are “Indians” holding Kenyan citizenship. Finally, the census form makes it possible to declare to be “Pakistani” or “Other Asians” (but it is known that the numbers of the latter two categories are on average always very low).

  • 41 Is this scenario possible? In principle, Kenyan law does not allow a person to hold two passports.

82It can be assumed that people who have retained their Indian or Pakistani nationality said they were “Indians” or “Pakistanis” (unless they tried to affirm their integration into the Kenyan society). In contrast, one can hardly imagine the options taken by people with dual nationality (Indian and Kenyan41); or those seeking to cling to their Indian origin, despite their Kenyan nationality (acquired recently perhaps), in order to preserve their cultural inheritance. Finally, let’s mention that even though the “Other Asians” category probably includes a majority of people from other regions of the Indian sub-continent, it absolutely does not guarantee that everyone falls under it (See Annex 1).

83Taking into account the quality of enumeration in 1989, it is sufficient to compare 1999 and 1989 data considering only relative figures. Contrary to relative stability – or slight increase – in the total number of “Indians” between the last two censuses, it is observed that the distribution of individuals based on the “tribes” variable changed distinctly during the inter-census period.

84The table shows that the weight of “Kenyan Asians” decreased by nearly 10 points between 1989 and 1999, whereas the proportion of other nationals, notably “Pakistanis”, increased significantly (by close to 50 % or more than 1,000 declared individuals in 1999). Does this change indicate a trend that results from a sustained migration current – but which went unnoticed due to lack of variation in the numbers between the two censuses? Or does it reflect a new “identity sensitivity” from a significant number of people who would have taken advantage of the 1999 census to affirm their Indian origin?

Table 3.13 Evolution of the variable Tribes between 1989 and 1999

Table 3.13 Evolution of the variable Tribes between 1989 and 1999

Sources: CBS

  • 42 The rumour (statistically unverifiable) had it that there were many Pakistanis among the recent imm (...)

85The census data cannot provide an answer to the second hypothesis. In contrast, the change in age structures according to the “tribes” variable shows that the first hypothesis needs serious consideration. Indeed, a look at the tables below shows that the ageing of among “Indian” and “Pakistani” groups is lower than ageing among “Asians” in general42. Women classified as “Other Asians” were observed to be younger.

Table 3.14 Evolution of the average age of segmented sub-populations according to the Tribes variable between 1989 and 1999 (in year)

Table 3.14 Evolution of the average age of segmented sub-populations according to the Tribes variable between 1989 and 1999 (in year)

Sources: CBS

86The “Kenyan Asians” category definitely represents a category whose average age increased most certainly during the last ten years (to a point that it weighed – due to its demographic volume – on the average age of the entire diaspora). This ageing can be explained by a little increase in the weight of young adults aged between 20 and 29 years old, an age group whose unique change in age structure between 1989 and 1999 we saw above (Graph 5). An examination of the “Indian” group’s age structure shows that the weight of the 0-19 year-olds shrank between 1989 and 1999 while that the 20-29 year-olds grew by 2 points (18.1 to 20.1 %). At the same time, the “Kenyan Asian” group only gained 0.2 points, growing from 15.7 to 15.9 %. Within the same age category, progression was still higher among “Pakistani” men and “Other Asian” women between 1989 and 1999. These changes are consequently in competition with ageing among “Kenyan Asians”. Moreover, between 30 and 39 years old, the proportion of “Kenyan Asians” decreased among both men and women, and increased among “Indian” men (see Annex 2). Such characteristics mitigate in favour of positive migratory balance between 20 and 29 years old, or even up to 40 years old. Even though we cannot rule out young adults leaving Kenya for other countries (this is indeed verified considering increased numbers of young people studying abroad), the number of those who choose to settle in the country therefore seems higher than those who left. Some of the newly arrived would therefore be “Indians” and not “Kenyan Asians”, which explains the decrease observed in the weight of the latter. The fact remains that the increase by about 10 points, in ten years, of the “Indians” category at the expense of “Kenyan Asians” could nevertheless be attributed to factors other than this migration. It would therefore be needed to turn to other hypotheses, particularly self-identification that is specific to every individual.

“Asian Kenyans” and “Indians”

87Further census data confirmed that “Kenyan Asians” and “Indians” (sticking to the two sub-communities that account for 90 % of the entire population classified in under the “Asians” group) had two distinct categories, each of them with a different migration history. Apart from the factors that have just been mentioned, the proportion of people born outside Kenya was thus twice higher in 1999 among “Indians” than among “Kenyan Asians” (60.6 % for “Indians” compared to 28.5 % for “Kenyan Asians”). This double “identity” link actually confirms that there is a correlation between stated ethnicity and citizenship.

Figure 3.7 proportion in age group of Kenyan Asians and Indians declaring to be born in Indian during the 1999 census (%)

Figure 3.7 proportion in age group of Kenyan Asians and Indians declaring to be born in Indian during the 1999 census (%)
  • 43 If the code Sudan actually corresponds to India – See Annex 1.

88Among various countries of birth of “Indians”, India is mentioned three out of four times43. 40.3 % of “Indians” were born in India as opposed to 12.6 % of “Kenyan Asians”.

89In every age category, those who described themselves as “Indians” were more often born in India than the so-called “Kenyan Asians” (graph above). For the latter, the more advanced their age was, the higher the likelihood of being born in India: three in ten of the oldest “Kenyan Asians” were born in India. In contrast, the younger the “Kenyan Asians” the less the frequency, particularly before the age of 20: more than 95 % of them having indeed been born in Kenya.

90For those who described themselves as “Indians”, the variation in ratio of those born in India follows a scale that is similar to that of “Kenyan Asians”. In contrast, these same proportions are higher whatever the age group. Between 40 and 55 % of “Indians” above 20 years of age were born in India. Between 20 and 50 years of age, the percentage was even higher, testifies which therefore of sustained migration as mentioned previously.

  • 44 Among “Pakistanis” and “Other Asians”, the percentages are even higher, 8.1 % and 12.0 %.

91Data on residence of people one year before the census shows further differences between “Indians” and “Kenyan Asians”. These differences are consistent with previous results on places of birth. Whereas the proportion of “Indians” who said they were living abroad in 1998 (one year before the census) was 7.7 %, this percentage did not exceed 4.3 % for “Kenyan Asians”44. Such data confirms a correlation between the self-determined ethnic identity and the period since arrival in the country. Among these “11th hour” migrants, 14.6 % of “Kenyan Asians” previously lived on the Indian sub-continent. But this proportion reached 49.8 % among new “Indian” residents, 53.6 % among “Pakistanis” and 58.4 % among “Other Asians”. The age distribution of these people also shows that most of the migrations take place from the age of 15 up to 45 years old, with the peak being attained earlier among the girls than among boys.

  • 45 See the chapter by Michel Adam on family and marriages.

92It can be observed that the curves dip very quickly beyond 25 years old. Considering the persistently tender age of marriage, one may think that at least a section of feminine immigration (and maybe to a lesser a section of masculine immigration as well) took place as a result of marriage proposals45.

Figure 3.8 Distribution per age of Indians declaring to be living in India one year before 1999

Figure 3.8 Distribution per age of Indians declaring to be living in India one year before 1999

Figure 3.9 Age of Indians declaring to be living in India during the 1999 census

Figure 3.9 Age of Indians declaring to be living in India during the 1999 census

Figure 3.10 Distribution per age of Indians declaring to be living in India in 1998, one year before the 1999 census

Figure 3.10 Distribution per age of Indians declaring to be living in India in 1998, one year before the 1999 census

Indians in Uganda

  • 46 See Calas (1991: 28-39).
  • 47 On August 4, 1972, Amin Dada decreed the expulsion of Indians within 90 days, claiming that God had (...)

93Uganda was a British protectorate from 1894 to 1962. The country then went through a tumultuous period, experiencing one inter-ethnic conflict after another, particularly under Idi Amin Dada’s dictatorial rule (1971-1979) and during the second Obote administration (1980- 1985)46. Many Ugandans endured military violence under the two regimes that were characterized by all sorts of political assassinations and massacres (the victims of atrocities between these two governments is estimated at ranging between 150,000 and 600,000). Targeted by both Amin Dada’s vindictiveness and a xenophobic decision to expel them, Indians were forced to leave the country in mass in 197247.

  • 48 Calas (1991: 29).
  • 49 Nabuguzi (1991: 134-140).

94As Indians were at the time playing a major role in manufacturing and business, their expulsion considerably weakened the country’s economy. Handed over to influential people in government, companies and businesses previously owned by members of the Indian diaspora were soon grinding to a halt as shortages, looting48, black market (magendo)49 and corruption became rife.

95In the early 1980s (under Obote’s government), several hundreds of Indians returned to Uganda with plans to reclaim their property, which had been sequestrated by a state-controlled organ (Departed Asians Properties Custodian Board or DAPCB). At the end of 1985, some of the Indians had to endure fresh violence in the looting that followed the fall of the Obote regime.

96It took the consolidation of Yoweri Museveni’s presidency (1988-89) for Indian presence in the country to be assured of guaranteed protection to enable them to progressively rebuild the former communities, which is only partly done currently.

97We mentioned in the introduction of this chapter that no statistics on the Indian diaspora had been released since the 1969 census. After numerous unsuccessful attempts, we managed to secure the Uganda Bureau of Statistics’ (UBOS) agreement to provide information we had requested. Although very incomplete, this data was part of the initial information released since the 1972 events. Before giving an account of this data we will, like we did for Tanzania and Kenya, delve into a description of the changes in the Indian population in Uganda during the period preceding independence.

Indian presence in Uganda up to 1972

98Various historical sources attest to the presence of Indian traders in Buganda and in the Great Lakes region by the 19th century. At the turn of the 20th century, the establishment of the British protectorate contributed to the arrival of a wave of traders and employees who were required to work in the colonial administration. Like in Tanzania and Kenya, the Indians’ level of education and their language skills, which enabled them to communicate in English, gave them some advantage over the Africans.

  • 50 Like other authors, Prunier (1990: 48, 58) brings out the poor social relations between Indians and (...)

99Organized in multiple communities, the immigrants established themselves in a society where ethnic relations between Africans were described as complex due to their political and cultural differences and the privileged support the British accorded to the Baganda people, an ethnic community living in central and southern parts of Uganda, along the shores of Lake Victoria50.

Table 3.15 Respective size of Asian, European and Africain populations living in Uganda (1914-1969)

Table 3.15 Respective size of Asian, European and Africain populations living in Uganda (1914-1969)

(*) Author’s estimates

Sources: UBOS (1991) and East African Statistical Department – Non-African population, 1948, 1959. Statistics on the Asians include people of Indian, Pakistani and Goan origin if the distinction was made.

100During the entire protectorate period, British presence remained low in Uganda, unlike in neighbouring Kenya where thousands had settled, particularly in the fertile areas of the highlands. The low European population, the distance from the sea, and lower urban development limited the number of Indian immigrants. Just before WWI, there were 3,651 Indians in Uganda (1914 census) while they were already three times more in Kenya. This gap (1 to 3) between the two countries remained the same in the years that followed, with Uganda having a population of 14,150 Indians in 1931 against 43,623 in Kenya the same year.

101Data from the censuses indicate an uninterrupted growth in the Indian population during the period between 1914 and 1969, reaching the maximum figure of 74,308 in 1969 (which was also a census year). In his article, Gérard Prunier suggests different estimates, exceeding 82,100 people in 1963, the year the country gained independence. Although the latter figure is not attributed to any source (in this regard, one can ask if these are revised estimates of the official figures released in 1959, considering undeclared immigration), Prunier explains the decrease in figures between 1963 and 1969 by the fact that some Indians had begun to leave the country, sensing unfavourable political changes. However, one may question the plausibility of this hypothesis, considering the supposed high growth in the figures between 1959 and 1963, just like the decline between 1963 and 1969 (7,800 people or 9.5 % of the 1963 figures).

Indians and British in Uganda

  • 51 In his article, Prunier also underlines this difference between Uganda and Kenya. He sees it as a p (...)
  • 52 Even though there are certain similarities between the Indian populations in the two countries, it (...)

102Until 1972, the population originating from the Indian subcontinent grew much more rapidly than the European population in Uganda – 91 % of which were British subjects in 1948. That same year (1948), there were over 10 “Indians” for every European in the country. In contrast, the Indian/European ratio in Kenya never went beyond 4 to 1 during the entire 20th century51. In Uganda, the growth of the Indian population was also faster than the growth of the African population. There were up to 11 “Indians” for every 1,000 African in 1959 (compared to 1 ‰ in 1914), ratios that are reminiscent of the situation observed in Kenya52.

103Finally, the British never established a large population in Uganda, with the highest population at about ten thousand recorded or estimated around 1960. The subjects of the Crown especially came to Uganda to serve in the protectorate administration. In 1948, less than 10 % were born there. Unlike the Indians (and even a good number of white settlers in Kenya), the British in Uganda were in the modern sense of the term “expatriate” civil servants who held short-term posts and were sure to soon return to Europe.

104A study of the age structures of Europeans and Indians confirms the distinct nature of these two sub-sets. The British were mostly above 25 years old and the 8-22 year-old age group were almost non-existent. They comprised couples without children or with young children. In contrast, the Indians had a pyramid age structure which indicates their situation of permanent immigrants who are bound to the country by retraceable family ties (see Figure 3.11 below).

Figure 3.11 Age structure of European and Indian populations living in Uganda in 1959

Figure 3.11 Age structure of European and Indian populations living in Uganda in 1959

Source: East African Statistical Department, Uganda Protectorate, Uganda Census 1959, Non-African Population. Pyramid on the left represents the European population and the pyramid on the right represent the Indian or Pakistani population (men on the left, women on the right). Photos by the author from an original booklet available at UBOS.

  • 53 Prunier (1990: 41) emphasizes the role played by Indians in Ugandan agriculture, another difference (...)

105The 1948 census gives indications of the economic activities of the Indian diaspora. With the exception of a small minority of singles (137 in number), women were absent in the work market. 50 % of men were engaged in wholesale or retail business, 29 % in manufacturing plants or maintenance and repair companies, 8 % were active in textile companies and 5 % were employed in the public service. Despite the economic importance of the investments they made, a very small number of families were involved in agriculture (2 % of the assets). This is the case of two famous families, the Madhvani and the Mehta, who were owners of big tracts of land known to be leading suppliers of sugarcane53.

  • 54 Varna: status rank that brings together a number of castes (jâti) among the Hindus. Third in the la (...)
  • 55 In August 2007, the Aga Khan and President Museveni jointly laid the foundation stone of the Bujaga (...)

106The 1948 census does not specify religion, with the exception of two main groups: Hindus (24 %) and Ismailis (17 %). The unusual mention of the Vaishya (19 %) in a census questionnaire obviously concerns the largest Hindu sub-group (therefore included in this group within which it is one of the main Varna54). The Ismailis are known to have played an important economic role in Uganda before 1972. Since then, thanks to the strength of businesses and Aga Khan Foundations in East Africa, the Ismaili community is still economically active in Uganda55.

Migration of Indians in Uganda and age structures

107Due to lack of data on this subject, it is not possible to know the number of Indian arrivals and departures in Uganda from the beginning of the 20th century until 1972. However, change in the age structures (showing ageing of little significance over these decades) tends to demonstrate that not all Indians finally settled in Uganda when they first migrated there. After staying in the country for a somewhat long period of time, some of them presumably left and were replaced by others.

108The relatively constant age structure is also the result of a high fertility rate and relatively low life expectancy. In the 1948 census, 50 to 54 year-old Indian women, born and living in Uganda said they had given birth to 6 children per woman (5.4 children per woman for those born abroad). The high fertility rate goes hand in hand with the high rate of marriages. In 1948, only 1 % of women aged over 20 years old (but not exceeding 25 years old) were still single, and only 3 % of them were widowed (marriage among widowed women is prohibited in principle among Hindus).

  • 56 On this subject, see the chapter by Michel Adam on family and marriages.

109The high numerical imbalance between men and women also contributes to maximal first-time marriage since, on the “marriage market”, men are numerically superior and much older. The table below shows that the sex ratio imbalance only began to reduce from 1959, except for the older age groups: women newly settled in Uganda arrived young at an age when they could embrace married life, and moreover were perhaps already sought as wives by their future families-in-law56. Table 3.16 shows slow ageing by age structure. Those over 40 year old were 10 % in 1931, 14 % in 1948, 15 % eleven years later, and reached 19 % late in 1969. In its 1959 census analysis, the East African Statistical Department observes moreover that, due to its “youth”, the Indian population was in a way uncharacteristic, compared to age structures of Indians in other countries under British influence. For example, the 35 % under 10 year-olds in Uganda in 1948 is only comparable to 26 % similar age group among Indians in Egypt in 1947; 14 % over 40 year-olds among Indians in Uganda comparable to 23 % similar age group among Indians in Egypt.

Table 3.16 Evolution of age structure and sex ratio of Asians between 1931 and 1969

Table 3.16 Evolution of age structure and sex ratio of Asians between 1931 and 1969

Source: Ugandan censuses. Sex ratio is calculated by dividing the number of men by the number of women for each age group (multiplied by 100)

Figure 3.12 Age structures of Indians living in Uganda in 1969

Figure 3.12 Age structures of Indians living in Uganda in 1969

Source: UBOS, Report on the 1969 Population Census – Each histogram represents 100 % of the men and 100 % of the women. The 60 and more age group is assimilated to a decanal group for the sake of calculation.

110The differences in age structures between Indians born in Uganda and expatriate Indians is another indication of constant Indian migration into the country. Whereas 42 % of them had been born in a country other than Uganda, “expatriate” Indians were until 1969 still characterized by their relative youth, thus a testimony of recent arrival in the country – until just before the 1972 expulsions.

  • 57 This is what makes the 1972 expulsions even more dramatic, as it significantly targeted Indians bor (...)

111Nevertheless, the number of Indians born in Uganda rose over the years with an inversely proportional frequency in age. In the 1969 census, close to 6 out of 10 Indians resident in Uganda had been born within the country but did not necessarily hold Ugandan nationality57. The graph below shows that beyond 20 years old, fewer and fewer women had been born in Uganda than men, most of them having arrived later in country. This is a visible phenomenon in the preceding pyramids. After 50 years old, higher male mortality destabilized the numerical relationship between men and women.

Figure 3.13 Number of Indian men and women born in Uganda according to their age in 1969

Figure 3.13 Number of Indian men and women born in Uganda according to their age in 1969

112In 1969, the 31,308 Indians not born in Uganda had come from India (69 %), but also a notable percentage from Kenya (20 %), and to a lesser extent from Tanzania (6 %). These statistics confirm migration between the three East African countries around this time, which we have already demonstrated.

Like in Kenya, Indian settlers in Uganda opted for towns

113Indians in Uganda are unequally distributed in the country’s four large regions. In 1959, 54 % lived in Buganda, 36 % in the eastern region, while only 4 % and 6 % lived in the country’s north and west respectively (Europeans were also remarkably distributed in the same manner).

  • 58 Prunier (1990: 44).

114Like in Tanzania and Kenya, Indians were concentrated in urban centres, the base of their economic activity. Gérard Prunier recalls that Indians of the emerging city of Kampala were in 1921 a third of the total population (954 out of 2821)58. In 1959, three big towns hosted half of the Indian community (49.7 %), namely Kampala (22,268), Jinja (8,883) and Mbale (4,575). The rest of the diaspora resided mainly in the three other big towns of Masaka, Soroti and Tororo. At the same time, 62 % of Indians lived in big towns with a non-African population of above 500 people. However, all Indian communities did not equally settle in urban areas. Thus 41 % of Goans resided in Kampala. This percentage fell to 31 % for people from other regions of India and 23 % for immigrants from Pakistan.

115The country’s first few years of independence did not change this geographical distribution. In 1969, 56 % of Indians lived in Buganda, three quarters of them being settled in the Kampala area, while 33 % lived in the eastern region, including a third in Jinja. 11 % of members of the diaspora lived in northern and western regions.

Image 3.5 Photograph of a Jinja street in 2005: Former Indian shops

Image 3.5 Photograph of a Jinja street in 2005: Former Indian shops

Source: Author’s photograph, January 2005.

  • 59 In 1969, it is likely that the rate of schooling among women aged over 30 years old was an approxim (...)

116The Indian population in Uganda also exhibited other characteristics of modern life. Indeed, like in Kenya, Indians were relatively well educated with a higher rate of schooling in 1969 than Indians in Kenya in 1999, among men and children aged over 10 years old and girls between 10 and 19 years old (see the table below 3.5). Even though there is no guarantee that the data is reliable, the high figures shows a level of education strongly higher than that of the African population, a situation that remained globally the same two decades later, as confirmed by the 1991 census data (infra)59.

Table 3.14 Proportion of the Indian population in Uganda according to levels of education in 1969

Table 3.14 Proportion of the Indian population in Uganda according to levels of education in 1969

Source: UBOS, 1969 Population Census.

Changes in the Indian Population since 1972

  • 60 Calas (1991: 29); Forrest (1999: 76-90).
  • 61 According to Forrest (1999), there were 85,000 to 90,000 Indians resident in Uganda in 1970, two ye (...)
  • 62 United Nations (1994).

117How many Indians were really expelled from Uganda by Amin Dada in 1972? Due to lack of statistical data, it is impossible to provide an accurate answer to this question. Historians nevertheless share the thinking that a large majority of members of the diaspora was forced to leave. Debates focus on the number (necessarily low) of families that managed to stay put (500 people according to Bernard Calas and 300 according to Tom Forrest)60. In addition, authors continue to discuss the number of Indians that were in the country in 1972: 50,000 according to Tom Forrest (in total contradiction with the census results and other figures published elsewhere by the author)61; up to 80,000 according to Edward Khiddu-Makubuya, considering that 6,000 people were left out in data collection during the previous census62.

  • 63 The 1991 census puts the number of Muslims at 39 %. Hindus having not been identified as such, it c (...)

118According to Tom Forrest, 1,000 Indians began to return to the country from 1989. He says that the returnees were mainly Ismailis, who before 1972 held privileged positions in the Ugandan economy (Forrest, 1999: 84)63.

Lessons from the 1991 Census

  • 64 We refer here to the only data that was sent to us by the Uganda Bureau of Statistics.

119The 1991 census data64 makes it possible to carry out a hindsight assessment of the scale of the expulsions as well as the number of returnees in the years that followed President Museveni’s ascent to power.

120The statistical documents provided by the UBOS indicate that there were 2,292 Indians in the file for individuals and 2,694 in the file for accommodation. This gap in number is a result of the fact that the UBOS did not include in its file for individuals people born in India who said they had an “ethnic origin” other than “Indian” or “Pakistani”. The file for accommodation brings a section of these people to light. Thus, considering that there were people left out in the census, it can be reasoned that the number of Indians in Uganda in 1991 was around 3,000.

  • 65 It should be recalled that the last census before the expulsion of Indians from Uganda (1969) showe (...)
  • 66 The difference between 624 and 583 raises interrogation but the census data is only “declaratory”.

121Out of the 2,292 individuals in the file for individuals, 925 (or 40 % of them) said they were Ugandan-born65. In addition, 624 people, 583 of them born in the country, said they had always lived in Uganda66. Analysis of the variable “length of stay (in the district)” for the other 1,668 Indians (2,292-624) who had not always resided in Uganda shows that in 1991 they had lived in the country for a relatively short time, less than three years for half of them. Even while assuming some of these people had previously moved from one place to another within Uganda, a study of these figures reveals that most of the enumerated people had recently arrived in the country. On the contrary, the number of people who that year (1991) had lived in the district for over 20 years (except for the already considered case of those born in the country) did not exceed 159. Were this data accurate, it would indicate that the number of Indians who remained in Uganda after the 1972 expulsions – and were still alive in 1991 – could not exceed 783 people (624+159). However, since the accuracy of what people told census takers cannot be ascertained, nor can the deaths of those who remained in Uganda, it is preferable to suggest as a figure an inclusive range of between 600 and 900. Thus, the previous estimates advanced by Bernard Calas (500) appear relatively consistent and close in any case to those arrived at in our own calculations. Moreover, it goes without saying that a debate on just a few hundreds of people elicits limited interest. Interpreting the results of the 1991 census only confirms in a more statistically rigorous manner the massive nature of the 1972 expulsions, which all observers had already indicated. Other publications could detail the conditions that enabled a tiny minority to stay in the country against all expectations.

“Returnees” or fresh migration?

  • 67 The 925 people who resided in India before settling in Uganda should not be confused with the 925 I (...)

122In his 1999 article, Tom Forrest contends that several hundreds Indians who were expelled in 1972 made their way back to Uganda, mostly from countries where they had found prior refuge, notably the United Kingdom, Canada, India and Kenya. Suffice it to say that the 1991 census left little room for assumption that there were returnees from Western countries. Immigrants then only made rare reference to the West. Thus only 19 of them said they had lived in Great Britain, and even less said they had lived in Canada (8 people). In contrast, 124 people said they had resided in Kenya. The vast majority of the immigrants said they had come from India and Pakistan, 92567 and 133 respectively. The latter accounted for 56 % of all the residents who lived before in countries other than Uganda.

Table 3.15 Distribution of Ugandan Indians in 1991 according to the place of their last residence and their number of years spent in Uganda (in the same district)

Table 3.15 Distribution of Ugandan Indians in 1991 according to the place of their last residence and their number of years spent in Uganda (in the same district)

Source: UBOS, Uganda Census, 1991

  • 68 Indian migrants in the 20-39 year-old age group who had been staying in Uganda for less than 10 yea (...)

123Considering the age of Indians who lived in Uganda in the 1990s, the assumption of massive return of the 1972 expellees is also ruled out. Those enumerated in 1991 are first and foremost characterized by their youth68. This is particularly the case of women: 82 % of women (compared to 71 % of men) were below 40 years old in 1991. Could these have been the returnees’ children? The hypothesis, if it has any basis, is only verifiable for some of the figures, seeing that 30 % of men and 45 % of women were not born in 1972. If there were returnees, they would appear among over 40 year-old immigrants, who represent less than a quarter of the Indians.

124With a masculinity ratio of 1: 6, the 1991 Indian population moreover appeared to be a “selected” population, which is typical of recently arrived migrant populations. Some families were in the processes of formation (with less than 300 under the age of 10 to 19 years old and 500 under 10 years old). However, the proportion of single men remained high. Thus 13 % of the men said they were “heads” of households while 47 % of them and 15 year-old youths were variously aggregated in composite households (including parents or non-parents), with some households having only male adults.

Figure 3.14 Age structure of Indians living in Uganda in 1991

Figure 3.14 Age structure of Indians living in Uganda in 1991

Source: UBOS, The 1991 Population and Housing Census

Urban base, and high social and professional positions

125Despite the 1972 events that separate two distinct periods in the history of the Indian diaspora in Uganda, its current geographical distribution in 1991 showed some similarities with the situation in 1969. The western and northern regions (6 % and 2 % respectively) attracted few “Indian” immigrants to settle, unlike the eastern region, close to the Kenyan border with an easier link to the capital and the Indian Ocean coast (29 % of the population). 63 % of the diaspora is now concentrated in the central region (as opposed to 56 % in 1969), with 49 % in Kampala district alone. Like in other East African countries, this distribution depends on the nature of activities they engage in. Only 1 % of the enumerated men said they were engaged in a profession related to agriculture (75 % of the entire Ugandan population). Professional activities, which are male-dominated within the diaspora (only 9 % of the activities were said to involve women, mainly in business activities), are mainly divided between business, manufacturing industry and liberal professions. A quarter of those enumerated (26 %) were engaged in various ways (bosses or employees) in small businesses, but more than a third held leading professional functions (18 % of the general managers, 19 % of liberal profession leaders), with only 17 % occupying intermediate positions (employees and technicians). There is even a lower proportion (14 %) for the manual professions (craftsmen and workers, including 5 % mechanics). This data, corroborated by information on the immigrants’ level of education, is testimony to a high degree of advanced qualifications. 77 % of men and 70 % of women in the above-16-year-old’ category have at least been to high school (senior secondary). Within this group, 36 % of men and 24 % of women attended university for one or several years. Moreover, the level of education among recent immigrants is higher than among those who have been in the country for long.

Table 3.19 Distribution of Indians in 1991 according to their level of education and the number of years spent in Uganda

Less than 10 years

10 to 20 years

21 years and more

None

7 %

3 %

20 %

Primary school level

21 %

31 %

19 %

Secondary school level

39 %

46 %

43 %

University level

33 %

20 %

18 %

Total

100 %

100 %

100 %

Source: UBOS, 1991 Census

From the 1991 data to those of 2001

  • 69 As already mentioned at the beginning of this chapter, the UBOS did not respond to our request to b (...)

126The 1991 census data are the most complete of all existing data on the Indian diaspora in Uganda – currently available data from the 2001 census only had people who explicitly said they were of “Asian” origin ( “Asians” in Uganda)69.

  • 70 A high estimate of 11,500 people would be acceptable. Forrest (1999) estimates that there were 9,00 (...)

127In 1991, 40 % of “Indians” in Uganda (including Ugandan citizens and foreigners) were born in the country. The 2001 census mentions 8,818 people of Indian origin who said they were not Ugandan nationals, 26.5 % of them born in Uganda (2,340 out of 8,818). It is not possible to reliably estimate the percentage of “Indians” holding Ugandan nationality while aligning the same percentage of people born in Uganda to that of 1991 (40 %). If this were the case, the entire the “Indian” community would be underestimated at about 2,000 people: (2,340+2,000)/ (8,818+2,000) =40 %. Considering this uncertainty – and considering the percentage of the people who were not enumerated – “Indian” population for 2001 can be estimated at around 10,000 people instead of 8,818 as announced by UBOS70.

  • 71 And unless we assume, and this would not be plausible, that there was a serious underestimation in  (...)
  • 72 Uganda Statistical Abstract, July 1996, “Tourist arrival by country of usual residence”, p. 24.

128Whether or not we retain the low estimates by UBOS71, this population is growing at a high rate, probably due in part to illegal immigration, considering the high number of tourist visas issued to Indian and Pakistani citizens. Thus during the period between 1990 and 1994, the number of tourist visas multiplied three and a half times for Indians (1,062 to 3,509) and five-fold for Pakistanis (140 to 721)72. Tourist visas issued to Chinese or Japanese citizens or nationals of other Asian countries also increased, but to a lesser degree (two or three-fold). According to Tom Forrest, these undeclared immigrants generally stay for a few months in Uganda and use the resources they mobilize during the brief stay to leave for other destinations, like North America (Forrest 1999). This assumption is, however, countered in part by the actual increase in the number of residents, as mentioned previously.

Indian diaspora in Uganda in 2001

  • 73 The UBOS sent us seven statistics tables that do not make it easy to specifically compare with the  (...)

129Apart from the numerical progression that has just been highlighted, available statistical data in 2001 hardly changes the broad picture of the Indian-Ugandan community from what can be seen in the 1991 census73. There is confirmation or strengthening of certain trends. This is the same case for the high urban population or the high level of social and economic position. One can note, however, among the new trends, the growing proportion of women engaged in professional activities.

130In the last decade (1991–2001), the average age among people of Indian origin generally stabilized. The growing migration maintained a demographic supply of youths that compensated the rate of ageing among the immigrants who were already in the country in 1991. In 2001, 3.4 % of members of the community were aged 60 years old and above, an age group that had decreased by 0.4 points compared to 1991. The average age of the entire diaspora rose to 27.3 years old in 1991; it is now 27.5 years old. The male population, which was slightly younger in 2001 (29.0 years old in 2001; 29.5 years old in 1991), probably testifies to the immigration of young bachelors. In contrast, women have been ageing by an average 1.5 years (25.3 years old in 2001; 23.8 years old in 1991), a change that was due to the decrease in number of women below 20 (45.1 % in 1991 compared to 38.5 % in 2001).

131The differential change in the two genders results in a slight rise in sex ratio within the under-10 and 15-19 year-old age groups. In contrast, the sex ratio within 25-34 years old experience a net decrease, implying ageing of the female population (notably among recent female immigrants). This data can be compared to the previous situation in which female immigrants were found. Should we conclude from this that the number of fiancées coming from the Indian sub-continent also significantly reduced, as claimed by informants? The under-5 year olds were fewer in 2001 than 10 years earlier due to the arrival of childless young migrants, who were yet to make a family in the host country. The family structures observed in 1991 and 2001 experienced little change. The percentage of married people only increased by two points, with more men than women staying longer as singles due to their superior numbers (63 % men within the over-20 year-old age group).

Figure 3.15 Sex ratio evolution according to age within the Ugandan Indian community in 1991 and 2001

Figure 3.15 Sex ratio evolution according to age within the Ugandan Indian community in 1991 and 2001

Source: UBOS, 1991 and 2001 censuses

Table 3.20 Marital status of men and women of more than 10 years old among the Uganda Indian community in 1991 and 2001

Table 3.20 Marital status of men and women of more than 10 years old among the Uganda Indian community in 1991 and 2001

Source: UBOS, 1991 and 2001 censuses

Living in Uganda and foreign place of birth

  • 74 This latter information was not provided to us by the UBOS.

132“Indians” resident in Uganda had, by 2001, became even more concentrated in urban areas. 69 % of them lived in the central region, with 59 % living in the capital district of Kampala. The neighbouring districts of Wakiso, Mukono and Jinja, which are close to Lake Victoria, compose, together with Kampala, an area where three out of four “Indians” settled. Curiously, Jinja town – which distinguished itself for its high Indian population – did not live up to its previous reputation (from 21.7 % of Indians resident in Uganda in 1959, it decreased to only 9.5 % in 2001). After taking possession of Indian businesses in 1972, Africans kept their hold on the town. This is the same situation in Mbale (a town close to the Kenyan border), historically the third largest settlement implantation for Indians in Uganda. It used to have 11.2 % of the Indian diaspora (1959), but now has just a small fraction of it (4.4 %). The UBOS review of the immigrants’ places of birth shows a change in conformity with the registration of new migrants or returnees. The percentage of those born in Uganda decreased and consequently that of people of Asian descent increased. However, the omission of people of Indian origin who said they were Ugandan nationals (and not of Asian origin) in the 2001 documents could only make this change more dramatic and make us more prudent in interpreting the data. Moreover, it should be borne in mind that every immigrants’ place of birth does not suggest the last place of residence before settlement in Uganda.74

Image 3.6 Photograph of an old private house in Mbale

Image 3.6 Photograph of an old private house in Mbale

Source: Author’s photograph, January 2005

133With the 2001 UBOS documents indicating 2,340 Uganda-born “Indians”, it is necessary to go back to the last inter-census period. Whereas 925 people were born in the country in 1991, this figure increased by 1,415 individuals within ten years, a change that does not take into account people who may have been left out of the census.

Table 3.21 Marital status of men and women of more than 10 years old among the Uganda Indian community in 1991 and 2001

Table 3.21 Marital status of men and women of more than 10 years old among the Uganda Indian community in 1991 and 2001

Source: UBOS, 1991 and 2001 censuses

The strengthening of socio-economic positions at the top of the hierarchy

134The distribution of professional activities among members of the diaspora did not change much between 1991 and 2001. There were more and more Indians holding positions of responsibility, with 43 % of them being either self-employed, managerial staff or holding management positions. The “Indians”, who accounted for 0.04 % of the country’s population (4 “Indians” out of 10,000 Ugandans), held 3.5 % of the management positions (830 out of 23,458 posts).

135In the last decade, there was an important change in women’s work: 16 % of women were engaged in professional activity as opposed to 9 % ten years earlier. This progression is especially remarkable considering that a third of these women hold positions of high responsibility while only 10 % of them are unskilled workers.

Table 3.22 Socio-professional structure of the Indian population of Uganda during the 1991 and 2001 censuses

Table 3.22 Socio-professional structure of the Indian population of Uganda during the 1991 and 2001 censuses

Source: UBOS, 1991 and 2001 censuses

136A slight increase was noted in the proportion of assets in agriculture, but the percentage of engagement in the sector remained insignificant in a country where this sector remains massively dominant, employing 83 % of the women and 71 % of active men in 2001 (total population figures increased slightly in comparison with 1991).

137In conformity with the nature of the activities they engage in, the level of education among “Indians” rose once again during the last inter-census period. From 39 % in 1991 (33 % men and 48 % women), the percentage of “Indians” who had not gone beyond primary school fell to 26 % in 2001 (22 % men and 33 % women). These figures can generally be compared with the 89 % of Ugandans who had not gone beyond the same level of schooling. On the other extreme of the academic spectrum, the percentage of “Indians” who had gone to university rose from 24 % in 1991 (29 % men and 15 % women) to 36 % in 2001 (40 % men and 31 % women). This progression was more in women’s favour. Moreover, with only 2.7 % of the African population having reached university, the contrast is generally sharp.

138More than in Tanzania and Kenya, the position of “Indians” in Uganda which otherwise portrays them as a leading minority – they are not in any way involved politically and they do not participate in the country’s government – but at least they participate at the socio-economic level, seeing as they are highly-educated, active and prosperous. Not only are the levels of activity in this community higher than the average of the country’s population and exceeds that of all other (African) communities, but a number of members of the diaspora also create activities of their own. Whereas their population is still seven times lower than during the period before 1972, “Indians” in Uganda are, it appears, on a rising trend that could propel them to the economic positions that they used to hold a half-century earlier.

Conclusion

  • 75 Bearing in mind that an insignificant part of these African minorities consists of refugees from ne (...)

139Census protocols are currently the most accurate and reliable largescale means that enables us to know population size and composition. Towards the middle of the last century, the Indian population reached a figure above 350,000 in the three East African countries following accurate enumeration initiated by the British administration. After a sharp fall during the years after the independence upheavals, the Indian diaspora partly reconstituted itself (about 150,000 people in 2001) – with the help of fresh immigrants from India, to some extent – but they never, in proportion with the African population, regained their numbers during the colonial era. The case of Uganda – where the numbers of “Indians” were little over a long period of time – is both exemplary and special, due to their near total expulsion in 1972. The “Indian” population in this country, which has been growing rapidly since late 1980s, recovered especially with the help of fresh immigrants – and not, as claimed frequently, due to the return of former immigrants. However, and with a likely figure of about 10,000 (in 2001), its relative size further reduced both in relation to the entire African population and other foreign minorities, like Sudanese (164,000), Rwandans (106,000), Congolese (73,000) and Kenyans (35,000)75. In the case of Tanzania, the population’s reconstitution at independence was probably a lot higher, although the total absence of statistical data since the end of the colonial era prevents any succinct conclusions in this respect. To obtain acceptable statistical findings, it would be needed to use methods of indirect estimation in the main cities of Tanzania, an undertaking that is beyond the means of this study. Finally, in Kenya, whose situation is by far the best known, the “Indian” community plays – perhaps more than before – a central economic role, despite a major decline in population (by about 50 %) since independence. Kenya, which is, like in the past, home to two thirds of the diaspora in the three countries, seems to have also become a passage route, for temporary immigrants, who are attracted, in the short or medium term, to more faraway destinations.

140The fate of people from the Indian sub-continent in East Africa is dependent upon multiple risks: individual and collective will to assimilate, political will of governments to take advantage of their presence while managing to keep xenophobic impulses at bay.

Bibliographie

Bibliography

ADAM, Michel 2010 (2006), « A Microcosmic Minority: The Indo-Kenyans of Nairobi », in Hélène CHARTON-BIGOT & Deyssi RODRIGUEZ-TORRES (eds), Nairobi Today. The Paradoxe of a Fragmented City. Dar es Salam, Mkuki na Nyota, Nairobi, IFRA: 215-268.

CALAS, Bernard

1991, « La violence et ses conséquences urbaines à Kampala », Politique Africaine, 42: 28-39.

1998, « Des contrastes spatiaux aux inégalités territoriales », in François GRIGNON & Gérard PRUNIER (eds), Le Kenya contemporain. Paris, Karthala: 13-51

COUPLAND, R. (1938) 1968, East Africa and its Invaders. Oxford, Clarendon Press.

FORREST, Tom 1999, « Le retour des Indiens en Ouganda », Politique africaine, 76: 76-90.

GOLAZ, Valérie 1997, « Les migrations internes au Kenya, 1979-1989 », Documents et manuels du CEPED, 6.

Population & Sociétés (INED) 1999, « Tous les pays du monde », 348.

MANGAT, J.S. 1969, A History of the Asians in East Africa, 1885-1945. New York, Oxford University Press, 216 p.

NABUGUZI, Emmanuel 1991, « Le magendo en Ouganda », Politique africaine, 42: 134-140.

O.N.U-U.N. 1994, The Culture of Violence. Paris, University Press, 292 p.

PRUNIER, Gérard 1990, L’Ouganda et la question indienne (1896-1972). Paris, Éditions Recherche sur les Civilisations, 256 p.

Annexes

Annex 1 Quality of data on Indian population from the last two Kenyan censuses

Information on people of Indian origin living in Kenya has been extracted directly from soft census files. For the 1989 “Asians” file, like that of 1999, we had requested the CBS to extract from the reference file of individuals who responded to question regarding one of the four details of the “tribe(Tribe or Nationality) variable, referenced as P14 in 1989 and P13 in 1999.

Table 3.23 Codes and terminology selected from the Tribe variable

Table 3.23 Codes and terminology selected from the Tribe variable

Source: CBS, Code list for 1989 and 1999 population and housing pilot census

Although the census was declarative in nature, it can be considered that there is a major difference between the four details presented in the table above. Those who said they were “Kenyan Asians” should first be considered Kenyan citizens from the Asian continent (and in fact essentially from the Indian sub-continent), while other details of “Tribe” variable first defined foreign nationals who had not acquired Kenyan nationality. The census forms have codes 46, 47, 48 for 1989, and codes 52, 53, 54 for 1999 called “Foreigners–Asians”, while a distinction is made between “Kenyan Asians” (Code 39), “Kenyan Europeans” (Code 40), “Kenyan Arabs” (Code 41) and “Other Kenyans” (Code 42), the same distinctions between Kenyan citizens from Kenya’s forty or so African ethnic groups.

Since the census is merely declarative, it is possible that “Asians” of Kenyan nationality declared in 46, 47 or 48, to show their attachment, for example, to the continent and attachment to their culture of origin, and vice versa that “Indians”, “Pakistanis” and “Other Asians”, depending on how long ago they settled in Kenya, said they were “Kenyan Asians”, although they were not holders of Kenyan nationality. We assume however that such cases are few.

a) Number of Entries in the 1989 Census Form (s)

The data on “Asians” extracted from the 1989 census were provided in two separate soft files. The first file contains the entire “Asian” population living in Kenya, with the exception of inhabitants of Nyanza Province. Indeed, for technical reasons, it was impossible for the CBS to have access to digitized individual entries from this province. In order to correct this gap, a replacement (second) file was generated by the CBS in a bid to recover data from the province on “Indians” drawn from the 5 % sample of the census. Thus, the first file has 82,905 entries compared to a mere 165 for the second. It can be agreed to multiply by 20 the latter figure in order to get an estimate of “Indian population” resident in Nyanza Province, about 3,300 inhabitants. The files therefore enable us to estimate at 86,205 in 1989 the total population size that we want to study, but we can reveal that this estimate differs from the one mentioned in the 1989 census released by the Central Office Statistics (published in March 1994) which should be considered as official.

The last enumeration according to “ethnic groups” (tribes)76 released by the Kenyan government in 1989 showed that there were 89,185 people belonging to the Asian community and who were distributed as follows: The gap of nearly 2,980 people between the two sources is not negligible (+3.46 % for official data compared to the estimated data from the two soft files) and can be explained by approximation of Nyanza inhabitants. One possibility is that this gap demonstrates misunderstanding of the definition of the “Asian” category. It may also be a compilation error. Moreover we know that some totals in the official publication77 are erroneous, for example those on page 6-46 of Kenya Population Census Volume I (1989).

Table 3.24 Official distribution of the Asian population during the Kenyan 1989 census

Code

Ethnic group (tribe)

39

Kenyan Asian

52 968

46

Indians

29 091

47

Pakistanis (Pakistanais)

1 862

48

Other Asians (AutresAsiatiques)

5 264

TOTAL

89 185

Source: CBS, Office of the Vice President Ministry of Planning and National Development, Kenya Population Census, Volume I, 1989, March 1994, Table 6: Population by tribe, sex and district.

Table 3.25 Distribution in the table at 5 % of “Asians” by “national group” and conversion for a total national scale workforce in Kenya

Value Label

Value

Frequency

Percent

Total Population

ASIAN-KENYAN

39

2675

0,25 %

53500

INDIANS

46

1464

0,14 %

29280

PAKISTANIS

47

91

0,01 %

1820

ASIANS

48

285

0,03 %

5700

Total Asian People

4515

90300

Total Kenyan People

1074131

21482620

Source: Statistical Confidentiality and the Construction of Anonymized Public Use Census Samples - a draft proposal for the Kenyan Microdata for 1989. Agnes A. Odinga and Robert McCaa, Minnesota Population Center, July 1, 2001, http://www.hist.umn.edu/​-rmccaa/​anon-kenya89-ver1.doc.

Although more inaccurate, another possible appreciation of the number of “Indians” living in Kenya in 1999 is available through the 5 % sample, which can be accessed through the Minnesota Population Center (Integrated Public Use Microdata Series – IPUMS – available on http://www.ipums.umn.edu/​). The numbers obtained this way (4515 x 20 = 90 300) are comparable (slightly higher) to those mentioned in the Kenyan statistics.

In conclusion, and despite the shortcomings in the two files mentioned below (82 905 + 3300), no better reliable and accurate sources exist to account for the 1989 population of Indians in Kenya. As the files sent by the CBS provide indications of the individual characteristics of every enumerated person, they made it possible to bring out the structural composition of the under-population as a whole, which is comparable to what was revealed by the subsequent census in 1999.

b) Number of Entries in the 1999 Census Form

Considering that no publication brought out the ethnic origins of the people enumerated in 1999, the 89,310 people identified under the “Asians” category in the unpublished file from the same census are the new statistical reference for this population. To account for it, we spread the “Asian” group between the four referenced “ethnic groups”.

Table 3.26 Distribution per national group of the Asian population living in Kenya in 1999

Code

Tribe (Ethnic group)

number

45

Kenyan Asian

44 461

52

Indians

35 980

53

Pakistanis

2 897

54

Other Asians

5 972

TOTAL

89 310

Source: CBS, 1999 census

For 1999, all the data on people of Indian origin appear in the same file, contrary to the 1989 presentation. However, this arrangement is not prejudiced by total absence of errors as we are soon going to demonstrate.

c) Coding Errors in the 1999 Census Forms

A study of the places of birth variable in 1999 brought out coding errors on the details of some variables. This is the case with mentioning the place of birth as it appears during comparative reading of the (joint) tables that follow:

Table 3.27 Distribution by Country of birth of Asian (as enumerated in Kenya in 1989 and 1999)

Table 3.27 Distribution by Country of birth of Asian (as enumerated in Kenya in 1989 and 1999)

The comparison between 1989 and 1999 needs to give priority to relative values because the Asians of 1989 enumerated here (82905) do not include the 3,300 inhabitants of Nyanza province.

Source: CBS – from the files of the 1989 and 1999 censuses

In 1999, it seems that no Indian was born in India (not even in any Asian country) while those who indicated they had been born in Sudan were 74.4 % of the “Asians” born abroad (and 25.1 % of the total number of “Asians” resident in Kenya). The unlikelihood of these numbers is again acknowledged in the results of the 1989 census which shows that 79.5 % of those born abroad were born in an Asian country while a minute percentage was born in Sudan. The most plausible explanation of this error is an unintentional confusion of Code “095” and Code “005”. Thus, all “Asians” born in India were in 1999 classified in the slot of those born in Sudan.

The same confusion could explain the total lack of figures in 1999 for the following codes: European countries (094), American countries (096) and Rest of the world (097). The insertion of these numbers in Somalia (004) and Other Africa (006) codes cannot be ruled out. There is nevertheless reason to wonder why the sum of Other Africa, American countries and Rest of the World in 1989 could not correspond to the figure for Other Africa alone in 1989 (11.1 %), a fact that waters down this theory.

Statistics on people from Uganda, Tanzania and Ethiopia do not seem suspect at first over the same type of errors. For immigrants from Uganda, it will be noted however that the sharp increase in their number between the two censuses also corresponds to a high immigration in Uganda.

Finally, let us observe that the “Other Asians” category in Tables A2 and A4 should also be treated with caution, because Kenyan statistics do not provide information on the nationality of these people of Asian origin and it would be improper to consider them as coming from the Indian sub-continent.

Annex 2 Structures per age, sex and ethnic group (tribes) of people of Indian origin living in Kenya in 1989 and 1999

Table 3.28a 1989

Table 3.28a 1989

Table 3.28b 1999

Table 3.28b 1999

Table 3.28c Difference 1999–1989

Table 3.28c Difference 1999–1989

Sources: CBS, Kenyan 1989 and 1999 censuses

Annex 3 Distribution of the Asian population in Kenya in 1999

Figure 3.16b

Figure 3.16b

Employees by district population living in Kenya in 1999 “Asians “

Annex 4 Official publications on population from the three countries studied

Tanzania

– Nations Unies, Département des questions sociales, Division de la population, Rapport sur la population des Territoires sous tutelle, n° 2, La Population du Tanganyika, Lake Success, New-York, 1/9/1949 (Document available at CEPED)

– Tanganyika Statistical Abstract, 1938-1951, The Government Printer, Dar es Salaam, 1953

– Tanganyika Statistical Abstract, 1958, The Government Printer, Dar es Salaam, 1958

– US Census Bureau, International Database, Tanzania 1957

– National Bureau of Statistics, 1967 Population Census, Volume 3, Demographic Statistics, 1971

– The United Republic of Tanzania, 1978 Population Census, Volume 1, Methodology Report

– Tanzania, 1988 Population Census, Preliminary Report, Bureau of Statistics, Dar es Salaam

– The United Republic of Tanzania, 2002 Population and Housing Census, General Report, Central Census Office, National Bureau of Statistics, President’s Office, Printed by Government Printer, Dar es Salaam, 2003

Kenya

– Colony and Protectorate of Kenya, Report on the Census on the Non-Native Population of Kenya Colony and Protectorate Taken on the Night of the 25th February 1948, Nairobi, 1953

– Colony and Protectorate of Kenya, Report on Census of Non-Native Employees, 1949, East African Statistical Department, 1950

– Republic of Kenya, Kenya Population Census, 1962, Volume IV: Non-African Population, Statistics Division, Ministry of Economic Planning and Development, March 1966

– Republic of Kenya, Kenya Population Census, 1962, Tables: Advance Report of Volume I & II, Economics and Statistics Division, Ministry of Finance and Economic Planning, March 1966 (IFRA document available at Nairobi)

– Republic of Kenya, 1969 Population Census, Volume I: Analytical Report, Central Bureau of Statistics (IFRA document available at Nairobi)

– Republic of Kenya, 1979 Population Census, Volume I (Table 2: Population by Sex, Tribe or National Group and District)

– Republic of Kenya, Kenya Population Census, 1989, Volume I, (Table 6: Population by Tribe, Sex and District), Central Bureau of Statistics, Office of the Vice President, Ministry of Planning and National Development, March 1994

– Republic of Kenya, 1999 Population and Housing Census, Volume I: Population Distribution by Administrative Areas and Urban Centres, Central Bureau of Statistics (CBS), Nairobi, Kenya, January 2001

– Republic of Kenya, 1999 Population and Housing Census, Volume II: Socio-Economic Profile on the Population, Central Bureau of Statistics (CBS), Nairobi, Kenya, January 2001

– Republic of Kenya, 1999 Population and Housing Census, Volume III: Analytical Report on Population Dynamics, Central Bureau of Statistics (CBS), Nairobi, Kenya, January 2001

– Republic of Kenya, 1999 Population and Housing Census, Volume XI: Analytical Report on Gender Dimensions, Central Bureau of Statistics (CBS), Nairobi, Kenya, August 2002

– Republic of Kenya, 1999 Population and Housing Census, Volume IV: Analytical Report on Fertility and Nuptiality, Central Bureau of Statistics (CBS), Nairobi, Kenya, August 2002

Uganda

– Uganda Bureau Of Statistics (UBOS), Report on the 1969 Population Census, Volume I: The Population of Administrative Areas, Nov. 1971; Volume II: The Administrative Report, April 1974.

– The Republic of Uganda, The 1991 Population and Housing Census, Analytical Report, Volume I, Demographic Characteristics, Statistics Department, Ministry of Finance and Economic Planning, Entebbe, Uganda, May 1995.

– The Republic of Uganda, The 1991 Population and Housing Census, Analytical Report, volume II, Socio-Economic Characteristics, Statistics Department, Ministry of Finance and Economic Planning, Entebbe, Uganda, May 1995.

– The Republic of Uganda, 2002 Uganda Population and Housing Census, Main Report, Uganda Bureau Of Statistics (UBOS), March 2005

– The Republic of Uganda, 1996, Statistical Abstract, Statistics Department, Ministry of Finance and Economic Planning, Entebbe, Uganda, July 1996

– Uganda Bureau of Statistics (UBOS), Migration and Tourism, Report IV (2000-2004), August 2005

– Uganda Bureau Of Statistics (UBOS), Statistical Abstract, 2005

– Uganda Districts, Information Handbook, Fountain Publishers, Kampala, Fifth edition, 2002

– US Census Bureau, International Database, Uganda, 1959, 1969, 1991 http://www.census.gov/​

–Uganda Protectorate, Uganda Census 1959, Non-African Population, East African Statistical Department, June 1960

Notes

1 The author sincerely thanks Kenya’s Central Bureau of Statistics and Uganda’s Bureau of Statistics for agreeing to release, for scientific purposes, information on Indian communities extracted from the last Kenyan and Ugandan censuses, when this data was in public circulation. Without the trust accorded by these two institutions, this work would not have seen the light of day. Similarly, many thanks go to Bernard Charlery who in January 2005 welcomed us to the French Institute for Research in Africa (IFRA) and whose help and support were invaluable.

2 Ugandan 1969 Census.

3 Whereas the terms “race”, “racial group”, “tribe” and “ethnic group” were more or less synonymous during the colonial era (and until the release of the first census in independent Tanzania in 1967), governments of English-speaking countries now use the terms “tribe” or “ethnic group” to refer to “ethnic community”.

4 Coupland (1938, 1968).

5 This figure of 6,000 is high compared to 3,000 enumerated in 1900 in Tanganyika alone. See infra.

6 Prunier (1990). See Annexe III of Prunier’s publication for a historical presentation of Tanzania’s Indian community.

7 M.F. Lofchie (1965). See the article by Marie-Aude Fouéré in this volume.

8 United Nations, Department of Social Affairs, Population Division, Trust Territories Population Report, no. 2, Population in Tanganyika, Lake Success, New York, 1 September 1949.

9 It should be noted that the redrawing of borders at the initiative of the Belgian and British colonialists between Tanganyika, Rwanda and Burundi complicated statistical follow-up.

10 Wikipedia, http://en.wikipedia.org/wiki/Tanzania, 15 March 2006.

11 The exercises “seem to have consisted in enumerating people in certain regions, complemented by assessments carried out in the other regions, in order to arrive at the total population for the entire Territory”, UN (1949: 23).

12 “In 1921, Germans were repatriated and none of them was allowed to come back to the Territory. European agriculture and trade had not yet shown the effects of the East African campaign, nor the depression that followed the war, UN (1949: 81).

13 Tanganyika Territory, 1932. Though it is not possible to access this publication in Tanzania, a copy is available at the Archives of Canada library under HA2131 A5 1931a. The British government produced similar documents, using the same methodology, for Kenya and Uganda. Various easily accessible sources via the Internet mention the 1931 census, see for example: Wikipedia, http://en.wikipedia.org/wiki/Tanzania

14 Thus, the Indian population is estimated to be 23,422 on pages 140 and 142 of the report, but at 32,706 on page 141. In addition to the 23,422 people of Indian descent there are 1,722 Goans (UNO-UN, 1949).

15 Remember that our assumption is that Arabs had been assimilated to the “Asians” category in 1931. If the Arabs had been omitted from the 1931 statistics, then the 32,398 “Asians” were exclusively Indians and Goans, and in this case, there was a high decrease in the number of Indians between 1931 and 1938 (from 32,398 to about 21,000). Can such a decrease be credible?

16 There is still no distinction between Pakistanis and Indians despite the 1947 separation.

17 Having attained independence in December 1963, Zanzibar united with independent Tanganyika in April 1964 in the follow-up of a political revolution that was hostile to the minority Omani Arabs. The new state became Tanzania, joining two names TANganyika (continental territory henceforth called “Mainland”) and ZANzibar (island territory comprising two main islands Unguja (often confused with the entire archipelago) and Pemba, and about ten smaller islands.

18 “(…) The February 1952 census focused on all non-African residents as well as Africans resident in registered urban centres and other areas with high population density”, Tanganyika Statistical Abstract, 1958.

19 The high influx of Europeans (mostly British) entering and leaving Tanganyika may also be noted. However, this movement is largely a result of the limited duration of contracts of civil servants posted to the colonies.

20 Even though, from a geographic point of view, it was part of the Zanzibar archipelago, Mafia Island was then politically linked to the mainland territory of Tanganyika and not to the political unit of Zanzibar, and therefore cannot be counted together with the rest.

21 This table does not fully reflect the cultural and religious diversity of Indian communities. Hindus and Muslims are divided into several denominations which constitute different sub-cultures.

22 These are ratios calculated using the 1957 (and 1967) census data.

23 High Commission of India (http://www.hcinairobi.co.ke/). Indian diplomatic officials generally say that they cannot numerate Indians who do not hold Indian passports. Web site: http://www.hcindiatz.org/indtanu.htm

24 In contrast, all published statistics are freely accessible. Various censuses conducted by the British before independence generally distinguished “natives” from “non-natives”, each of these two categories being subdivided statistically into “ethnic” (tribal) or national sub-categories. Not all African countries have this knowledge, as some states carried out the first census of their population only shortly after independence. As we have mentioned above, we were kindly provided upon request with the digital data on 1989 and 1999 Kenyan censuses by the Census Office Statistics of Nairobi (CBS). At the end of the chapters (Appendix 1) there is an assessment of the quality of data on the Indian population from these two most recent censuses in Kenya.

25 For a more complete history in French, see Prunier (1990).

26 Restricting the issuance of the lion’s share of business licences to Kenyan citizens.

27 It has not been possible to update this article by taking into consideration the last census.

28 The remaining 40 % are “Asians” who had moved from one district to another within Kenya.

29 A figure between 3 and 8 % seems more realistic.

30 Supposing, in a totally arbitrary manner, that one entry out of four is illegal, leading within five years to either to status regularization or to relocation in another country, there would be less than 10,000 Indians currently illegally resident in Kenya.

31 Particularly Michel Adam, who advances a figure of 120,000 for the year 2002 based on approximate enumerations with the help of local authorities (Adam 2006: 302).

32 In its publication Population and Housing Census, Analytical Report on population Dynamics, vol. III, 2001, p. 48, the Central Bureau of Statistics estimates the number of “Asian” immigrants without Kenyan nationality (foreigners) living in Kenya in 1999 at about 82,000. This figure contradicts data from the same institution in Table 3.26 of Annex 1.

33 Locations in Kenya are administrative units that are equivalent to the French “communes”. Nairobi is an exception as each location is a “residential estate”.

34 This actually refers to Nairobi West estate, which is part of Kibera Division. For Nairobi residents, Kibera particularly refers to the sprawling slum situated north of the division with the same name.

35 Remember that we identified the Indian population living in Kenya using the P13-Tribe census variable. Four modes of this “tribe” variable are used to define the population: “Kenyan-Asians”, “Indians”, “Pakistanis” and “Other Asians”. The first two modes account for 90 % of the individuals (Ref. Annex 1 data). We assume that there is a bias in saying that one is “Kenyan-Asian” or “Indian”: 13 % of “Kenyan-Asians” are born in India compared to 40 % of “Indians”.

36 Source: INED (1999). It should be noted that in the “Demographic and Health Survey”, the cyclical fertility indicator was estimated at 4.7 children for every woman in Kenya in 1998 (3.1 in the urban centres and 5.2 in rural areas). This figure was cited by INED in its publication mentioned above. However, the census results point to a slight underestimation of “momentary fertility” compared to the 1998 survey on population and health. See KENYA 1999, Population and Housing Census, Analytical Report on Population Dynamics – vol. III, pp. 13-21.

37 “Heads” meaning “heads of family”.

38 For the rates in Kenya, we calculate the figures using data from: KENYA 1999, Population and Housing Census, Socio-economic profile of the population – vol. II, Tables 1 and 2 on education.

39 It is worth noting with some surprise that, among the Asians, more boys and girls between 6 to 9 years old (and 10-14 year-old boys) have left school than Kenyan children, but it is in the Kenyan population that the highest number of 6-9 year-old children have never been to school.

40 During the census, the question asked was: “What is this person tribe?” It was recommended to the census surveyors that the people should be allowed to answer as they wish: “For Kenyan tribes, code using the tribe code list. Accept the answer as given to you without question. For Kenyans of other origins use the codes are provided. For example, persons originating from Asia should be coded ‘46’”.

41 Is this scenario possible? In principle, Kenyan law does not allow a person to hold two passports.

42 The rumour (statistically unverifiable) had it that there were many Pakistanis among the recent immigrants or “rockets”, some of them being clandestine. The same is true of Southern Indians, Sri Lankans, etc.

43 If the code Sudan actually corresponds to India – See Annex 1.

44 Among “Pakistanis” and “Other Asians”, the percentages are even higher, 8.1 % and 12.0 %.

45 See the chapter by Michel Adam on family and marriages.

46 See Calas (1991: 28-39).

47 On August 4, 1972, Amin Dada decreed the expulsion of Indians within 90 days, claiming that God had ordered him to do it. The series of events have been detailed in Gérard Prunier (1990).

48 Calas (1991: 29).

49 Nabuguzi (1991: 134-140).

50 Like other authors, Prunier (1990: 48, 58) brings out the poor social relations between Indians and Africans. He also mentions the complex relations between the British and the Indians, which resulted from both complementarity and rivalry.

51 In his article, Prunier also underlines this difference between Uganda and Kenya. He sees it as a possible explanation of the British settlers’ hostility towards Indians in Kenya.

52 Even though there are certain similarities between the Indian populations in the two countries, it is quite different when it comes to the Arab population. In Uganda, Arabs have always been a tiny minority; they were only 515 in 1931 and 1946 in 1959.

53 Prunier (1990: 41) emphasizes the role played by Indians in Ugandan agriculture, another difference when compared to Indians in Kenya. However, lit is noteable that even in Uganda, only a small minority of Indians were involved in agricultural activities as employers (engaging African workers), but never as independent producers and not in the least as salaried employees. See the introductory chapter of this volume.

54 Varna: status rank that brings together a number of castes (jâti) among the Hindus. Third in the labour division hierarchy (after the Brahmans and the Kshatrya), the Varna of Vaishyas represents the category of traders and farmers, who are the majority in East Africa. Hindus do not usually talk about their Varna status, but only about their caste. It is therefore surprising that the British administration used this in a document intended to be a population record. For a more detailed explanation of the caste issue, turn also to the introductory chapter of this collection.

55 In August 2007, the Aga Khan and President Museveni jointly laid the foundation stone of the Bujagali-Naminya hydroelectric power station, a project funded by the Aga Khan Economic Development Fund and the World Bank.

56 On this subject, see the chapter by Michel Adam on family and marriages.

57 This is what makes the 1972 expulsions even more dramatic, as it significantly targeted Indians born in the country who had more often than not lost ties with the country of their ancestors.

58 Prunier (1990: 44).

59 In 1969, it is likely that the rate of schooling among women aged over 30 years old was an approximate assessment, since the figures were established based on the declarations of the enumerated people and data compiled by the Uganda Bureau of Statistics (UBOS).

60 Calas (1991: 29); Forrest (1999: 76-90).

61 According to Forrest (1999), there were 85,000 to 90,000 Indians resident in Uganda in 1970, two years before the expulsions. If 50,000 of them were expelled, no more than 50 families remained in the country after the events as claimed by the author. On the contrary, if this figure of 50 families were retained, it would indicate that the number of those expelled would exceed 50,000 by far.

62 United Nations (1994).

63 The 1991 census puts the number of Muslims at 39 %. Hindus having not been identified as such, it can be assumed that they appear in the “Other religions” category which comprises 44 % of Indians (variable P05 of the census).

64 We refer here to the only data that was sent to us by the Uganda Bureau of Statistics.

65 It should be recalled that the last census before the expulsion of Indians from Uganda (1969) showed that a third of the population, about 26,000 people, had been born in the country.

66 The difference between 624 and 583 raises interrogation but the census data is only “declaratory”.

67 The 925 people who resided in India before settling in Uganda should not be confused with the 925 Indians who previously said they had been born in Uganda (numerical coincidence). Among the 925 people born in Uganda, 292 had resided in the country for less than 21 years (32 %). These could therefore be the returnees mentioned by Forrest (1999).

68 Indian migrants in the 20-39 year-old age group who had been staying in Uganda for less than 10 years (those who arrived between 1982 and 1991) were 53 % in 1981, while those who arrived in the previous decade were 43 %, and those who had been resident in the country for 21 years or more were 33 %. If we calculate the same percentages without including those under 20 years old (since they were not yet born 20 years ago), the recently arrived migrants tend to be even more significantly young: 70 %, 51 % and 33 % respectively between 20 and 39 years old during the mentioned periods of time.

69 As already mentioned at the beginning of this chapter, the UBOS did not respond to our request to be sent digitalized “individual” data from the 2001 census whose release was delayed to 2005. The enumerated people indeed had two options: ethnic group or citizenship. There was a danger of double confusion in this procedure: 1) between the two terms stated above (Ugandan citizens of “Indian” descent then failing to mention their ancestry); 2) between ethnic group in its limited sense ( “Asian”, according to popular reference in East Africa, more often than not means Indian or Pakistani) and ethnic group in the wider sense (Asian, which includes other parts of the sub-continent: Bangladesh, Sri Lanka, etc., or even other Asian countries: China, Japan, etc.). See UBOS, 2002 Uganda population and housing census, Main report, 120 p., March 2005.

70 A high estimate of 11,500 people would be acceptable. Forrest (1999) estimates that there were 9,000 “Indians” in Uganda in 1999.

71 And unless we assume, and this would not be plausible, that there was a serious underestimation in 1991.

72 Uganda Statistical Abstract, July 1996, “Tourist arrival by country of usual residence”, p. 24.

73 The UBOS sent us seven statistics tables that do not make it easy to specifically compare with the 1991 census results. The seven tables contain age, marital status, level of education, type of employment, religion, district of residence and country of birth statistics. The data on religion are unclear since the questionnaire does not explicitly indicate whether one belongs to the “African religions”, “Muslim” or “Christian” category. Hindus appear nowhere as such, but are (as it would appear) included in a final category entitled “other non-Christian religions”. As opposed to the situation in 1991, Indian Muslims were in 2001 more numerous than adherents to other religions: 34.2 % compared to 30.3 % in 1991.

74 This latter information was not provided to us by the UBOS.

75 Bearing in mind that an insignificant part of these African minorities consists of refugees from neighbouring regions affected by civil wars.

76 “Last” in the sense that the 1999 data based on ethnic groups (tribes) has not been released and we will be presenting it in this article for the first time with the kind assistance of the Permanent Secretary in the Ministry of Planning and National Development.

77 For an appreciation of the quality of the1989 Kenyan census, see: Golaz, 1997.

Table des illustrations

Titre Image 3.1 Photograph of an Indian owned shop in Tabora during German rule
Crédits Source: http://www.postcardman.net/​141727.jpg - Deutsch Ost-Afrika - Tabora - Indian shop.
URL http://books.openedition.org/africae/docannexe/image/937/img-1.jpg
Fichier image/jpeg, 46k
URL http://books.openedition.org/africae/docannexe/image/937/img-2.jpg
Fichier image/jpeg, 15k
Titre Table 3.3 Non-African male population between 20 and 49 years old by sectors of activity in Tanganyika – 1931 census
Crédits Source: Report on the Non-Native, Census 1931, September 1949.
URL http://books.openedition.org/africae/docannexe/image/937/img-3.jpg
Fichier image/jpeg, 40k
Titre Table 3.4 Entries and Exits in Tanganyika between 1938 and 1957
Crédits Source: East African Statistical Department (from records of Immigration and Passport office). Cf. from 1938 to 1948, Goans are included in the Indian population.
URL http://books.openedition.org/africae/docannexe/image/937/img-4.jpg
Fichier image/jpeg, 38k
Titre Table 3.6 Evolution of the population of Indian origin between 1931 and 1967
Crédits Sources: Censuses of Tanganyika and Tanzania
URL http://books.openedition.org/africae/docannexe/image/937/img-5.jpg
Fichier image/jpeg, 54k
Titre Figure 3.1
URL http://books.openedition.org/africae/docannexe/image/937/img-6.jpg
Fichier image/jpeg, 25k
Titre Image 3.2 Photograph of the Mombasa–Lake Victoria railway under construction
Crédits Source: Museum of Nairobi, temporary exhibition on Indian heritage (January 1995) Some Indians in the middle of the photograph supervise the work of Africans.
URL http://books.openedition.org/africae/docannexe/image/937/img-7.jpg
Fichier image/jpeg, 48k
Titre Figure 3.2 Evolution of the Indian population or the population of Indian origin living in Kenya (1840–1999)
Crédits Sources: CBS, Kenyan population census
URL http://books.openedition.org/africae/docannexe/image/937/img-8.jpg
Fichier image/jpeg, 22k
Titre Table 3.8 Evolution of the three main non-African populations living in Kenya between 1911 and 1989
Crédits Sources: The East African Statistical Department until 1948 then the Statistics Division.
URL http://books.openedition.org/africae/docannexe/image/937/img-9.jpg
Fichier image/jpeg, 16k
Titre Image 3.3 Maps of Kenyan districts, total population and Asian population per district
Crédits Source: UBOS – Kenyan 1999 Census Cartogrammes were made with the help of Dominique Andrieu (Tours, France)
URL http://books.openedition.org/africae/docannexe/image/937/img-10.jpg
Fichier image/jpeg, 28k
Titre Image 3.4 Map of Nairobi district – Indian Population by « location » (1999)
Crédits Source: CBS - Kenya 1999 Census
URL http://books.openedition.org/africae/docannexe/image/937/img-11.jpg
Fichier image/jpeg, 35k
Titre Table 3.9 « Locations » of Nairobi district having more than 2000 Asian inhabitants –Socio-economic indicators.
Crédits Source: CBS - Kenya 1999 Census
URL http://books.openedition.org/africae/docannexe/image/937/img-12.jpg
Fichier image/jpeg, 60k
Titre Figure 3.3 Birth place of people of Indian origin living in Kenya in 1999
URL http://books.openedition.org/africae/docannexe/image/937/img-13.jpg
Fichier image/jpeg, 33k
Titre Table 3.10 Birth place of Asians Living in Nairobi, Mombasa, Eldoret, Kisumu and Nakuru.
Crédits Source: CBS - Kenya 1999 Census
URL http://books.openedition.org/africae/docannexe/image/937/img-14.jpg
Fichier image/jpeg, 58k
Titre Table 3.11 Relationship with the referent person (household head) according to gender
Crédits Source: CBS – Kenya 1999 Census
URL http://books.openedition.org/africae/docannexe/image/937/img-15.jpg
Fichier image/jpeg, 36k
Titre Figure 3.4 Age structure of the population of Indian origin in Kenya in 1999
URL http://books.openedition.org/africae/docannexe/image/937/img-16.jpg
Fichier image/jpeg, 27k
Titre Figure 3.5 Age pyramid of the Kenyan population in 1999
URL http://books.openedition.org/africae/docannexe/image/937/img-17.jpg
Fichier image/jpeg, 28k
Titre Table 3.12a Proportion of the Indian population in Kenya according to education status in 1999
URL http://books.openedition.org/africae/docannexe/image/937/img-18.jpg
Fichier image/jpeg, 32k
Titre Table 3.12b Proportion of the urban Kenyan population according to education status in 1999
Crédits Source: CBS - Kenya 1999 Census
URL http://books.openedition.org/africae/docannexe/image/937/img-19.jpg
Fichier image/jpeg, 32k
Titre Figure 3.6 Variation in the weight of age group between 1989 and 1999
URL http://books.openedition.org/africae/docannexe/image/937/img-20.jpg
Fichier image/jpeg, 22k
Titre Table 3.13 Evolution of the variable Tribes between 1989 and 1999
Crédits Sources: CBS
URL http://books.openedition.org/africae/docannexe/image/937/img-21.jpg
Fichier image/jpeg, 26k
Titre Table 3.14 Evolution of the average age of segmented sub-populations according to the Tribes variable between 1989 and 1999 (in year)
Crédits Sources: CBS
URL http://books.openedition.org/africae/docannexe/image/937/img-22.jpg
Fichier image/jpeg, 13k
Titre Figure 3.7 proportion in age group of Kenyan Asians and Indians declaring to be born in Indian during the 1999 census (%)
URL http://books.openedition.org/africae/docannexe/image/937/img-23.jpg
Fichier image/jpeg, 39k
Titre Figure 3.8 Distribution per age of Indians declaring to be living in India one year before 1999
URL http://books.openedition.org/africae/docannexe/image/937/img-24.jpg
Fichier image/jpeg, 29k
Titre Figure 3.9 Age of Indians declaring to be living in India during the 1999 census
URL http://books.openedition.org/africae/docannexe/image/937/img-25.jpg
Fichier image/jpeg, 28k
Titre Figure 3.10 Distribution per age of Indians declaring to be living in India in 1998, one year before the 1999 census
URL http://books.openedition.org/africae/docannexe/image/937/img-26.jpg
Fichier image/jpeg, 27k
Titre Table 3.15 Respective size of Asian, European and Africain populations living in Uganda (1914-1969)
Légende (*) Author’s estimates
Crédits Sources: UBOS (1991) and East African Statistical Department – Non-African population, 1948, 1959. Statistics on the Asians include people of Indian, Pakistani and Goan origin if the distinction was made.
URL http://books.openedition.org/africae/docannexe/image/937/img-27.jpg
Fichier image/jpeg, 27k
Titre Figure 3.11 Age structure of European and Indian populations living in Uganda in 1959
Crédits Source: East African Statistical Department, Uganda Protectorate, Uganda Census 1959, Non-African Population. Pyramid on the left represents the European population and the pyramid on the right represent the Indian or Pakistani population (men on the left, women on the right). Photos by the author from an original booklet available at UBOS.
URL http://books.openedition.org/africae/docannexe/image/937/img-28.jpg
Fichier image/jpeg, 24k
Titre Table 3.16 Evolution of age structure and sex ratio of Asians between 1931 and 1969
Crédits Source: Ugandan censuses. Sex ratio is calculated by dividing the number of men by the number of women for each age group (multiplied by 100)
URL http://books.openedition.org/africae/docannexe/image/937/img-29.jpg
Fichier image/jpeg, 32k
Titre Figure 3.12 Age structures of Indians living in Uganda in 1969
Crédits Source: UBOS, Report on the 1969 Population Census – Each histogram represents 100 % of the men and 100 % of the women. The 60 and more age group is assimilated to a decanal group for the sake of calculation.
URL http://books.openedition.org/africae/docannexe/image/937/img-30.jpg
Fichier image/jpeg, 23k
Titre Figure 3.13 Number of Indian men and women born in Uganda according to their age in 1969
URL http://books.openedition.org/africae/docannexe/image/937/img-31.jpg
Fichier image/jpeg, 33k
Titre Image 3.5 Photograph of a Jinja street in 2005: Former Indian shops
Crédits Source: Author’s photograph, January 2005.
URL http://books.openedition.org/africae/docannexe/image/937/img-32.jpg
Fichier image/jpeg, 37k
Titre Table 3.14 Proportion of the Indian population in Uganda according to levels of education in 1969
Crédits Source: UBOS, 1969 Population Census.
URL http://books.openedition.org/africae/docannexe/image/937/img-33.jpg
Fichier image/jpeg, 32k
Titre Table 3.15 Distribution of Ugandan Indians in 1991 according to the place of their last residence and their number of years spent in Uganda (in the same district)
Crédits Source: UBOS, Uganda Census, 1991
URL http://books.openedition.org/africae/docannexe/image/937/img-34.jpg
Fichier image/jpeg, 60k
Titre Figure 3.14 Age structure of Indians living in Uganda in 1991
Crédits Source: UBOS, The 1991 Population and Housing Census
URL http://books.openedition.org/africae/docannexe/image/937/img-35.jpg
Fichier image/jpeg, 23k
Titre Figure 3.15 Sex ratio evolution according to age within the Ugandan Indian community in 1991 and 2001
Crédits Source: UBOS, 1991 and 2001 censuses
URL http://books.openedition.org/africae/docannexe/image/937/img-36.jpg
Fichier image/jpeg, 28k
Titre Table 3.20 Marital status of men and women of more than 10 years old among the Uganda Indian community in 1991 and 2001
Crédits Source: UBOS, 1991 and 2001 censuses
URL http://books.openedition.org/africae/docannexe/image/937/img-37.jpg
Fichier image/jpeg, 20k
Titre Image 3.6 Photograph of an old private house in Mbale
Crédits Source: Author’s photograph, January 2005
URL http://books.openedition.org/africae/docannexe/image/937/img-38.jpg
Fichier image/jpeg, 23k
Titre Table 3.21 Marital status of men and women of more than 10 years old among the Uganda Indian community in 1991 and 2001
Crédits Source: UBOS, 1991 and 2001 censuses
URL http://books.openedition.org/africae/docannexe/image/937/img-39.jpg
Fichier image/jpeg, 20k
Titre Table 3.22 Socio-professional structure of the Indian population of Uganda during the 1991 and 2001 censuses
Crédits Source: UBOS, 1991 and 2001 censuses
URL http://books.openedition.org/africae/docannexe/image/937/img-40.jpg
Fichier image/jpeg, 53k
Titre Table 3.23 Codes and terminology selected from the Tribe variable
Crédits Source: CBS, Code list for 1989 and 1999 population and housing pilot census
URL http://books.openedition.org/africae/docannexe/image/937/img-41.jpg
Fichier image/jpeg, 19k
Titre Table 3.27 Distribution by Country of birth of Asian (as enumerated in Kenya in 1989 and 1999)
Légende The comparison between 1989 and 1999 needs to give priority to relative values because the Asians of 1989 enumerated here (82905) do not include the 3,300 inhabitants of Nyanza province.
Crédits Source: CBS – from the files of the 1989 and 1999 censuses
URL http://books.openedition.org/africae/docannexe/image/937/img-42.jpg
Fichier image/jpeg, 65k
Titre Table 3.28a 1989
URL http://books.openedition.org/africae/docannexe/image/937/img-43.jpg
Fichier image/jpeg, 24k
Titre Table 3.28b 1999
URL http://books.openedition.org/africae/docannexe/image/937/img-44.jpg
Fichier image/jpeg, 24k
Titre Table 3.28c Difference 1999–1989
Crédits Sources: CBS, Kenyan 1989 and 1999 censuses
URL http://books.openedition.org/africae/docannexe/image/937/img-45.jpg
Fichier image/jpeg, 22k
Titre Figure 3.16a
URL http://books.openedition.org/africae/docannexe/image/937/img-46.jpg
Fichier image/jpeg, 26k
Titre Figure 3.16b
Légende Employees by district population living in Kenya in 1999 “Asians “
URL http://books.openedition.org/africae/docannexe/image/937/img-47.jpg
Fichier image/jpeg, 21k

© Africae, 2015

Conditions d’utilisation : http://www.openedition.org/6540

Acheter

Volume papier

amazon.fr
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search