Zambezia (1999), XXVI (ii).DEATH COLONIZED: HISTORICAL ADULT MORTALITYIN* RHODESIA (ZIMBABWE)RAVAI MARINDOUniversity of Zimbabwe, Centre for Population StudiesHistory is a nightmare from which I am forever trying to awake (JamesJoyce, Ulysses)AbstractUsing historical census reports from Rhodesia (now Zimbabwe) from 1901to 1962, this article discusses some of the socio-political forces that influencedthe collection of census data during the colonial period, ft evaluates theunderlying attitudes in which race, colonial interests and apartheid definethe quality and quantity of mortality data collected, and how in turn mortalitydata were 'colonized' to render a certain view of reality. The article concludesthat although the data used are historical, current African demographic datacollection has been influenced by these same historically rooted forces. Weemphasize the need for a critical evaluation of current demographic dataand research, the categories that we use in data collection, and assumptionsbehind them. These and other critical factors affect data collection by bothinternational organisations and national offices. The article further showsthat mortality differed between the two races and suggests that in countriesthat have experienced apartheid, there is need for the post-apartheid era todeal with the issue of collecting statistics by race. Data should be disaggregatedby race, but checks should be introduced to avoid repeated exploitation ofrace.INTRODUCTIONDemography has developed as a very quantitative discipline. Underlyingthis tradition is the assumption that demographic figures and statisticsare objective representations. Another common assumption is thatdemographic figures published by central, national and internationalstatistical offices may be technically unreliable but were basically collectedfor sound administrative purposes and the objective analysis of thesocial and economic situation of a population.Censuses have been used as tools for controlling populations. As farback as the biblical times, the Roman Empire carried out periodic censusesto identify people for taxation purposes. In fact, the Webster EncyclopedicUnabridged Dictionary of the English Language (1983) indicates that inAncient Rome the word census referred to 'the registration of citizens145146 HISTORICAL ADULT MORTALITY IN RHODESIA (ZIMBABWE)and their property for the purposes of taxation'. This basic character ofthe census has not changed.During the period of colonialism in Africa, censuses became tools notonly for taxation purposes but for segregating populations, and for creatingand emphasizing differences among the settler population and the nativepopulations. Census data became more than administrative output but inaddition a tool for defining and redefining categories that perpetuatedprevailing notions of social organization. In this sense, historical databased on censuses must be looked at as more than just output results butas indicators that defined and confirmed certain notions of political andsocial change.Historically, censuses from British colonies in Africa reflect policiesfor the harnessing of local labour, policies for control of movement ofboth the local and settler population, policies of deliberate silence andexclusion of part of the population and polices of racial superiority andapartheid.This article examines colonial censuses in Rhodesia, discussing andevaluating the way in which census data were collected and some of theunderlying motives for collecting these data. The article begins with abrief history of Rhodesia then moves on to use census reports forquotations and data that explain the socio-political context under whichthe census data were collected. Data on mortality and cause of death areused, where appropriate to show the excesses, omissions and gaps asevidence of how the data were 'colonized' that is, dominated for colonialpurposes to give a certain social reality. Colonization has a special meaningin this case which refers to how data are omitted, selectively collected, orover-collected, deliberately missed and sometimes distorted by agovernment in order to justify a certain way of allocating resources.DATA SOURCESMaterial for this article comes from the published census reports datingfrom 1891 to 1962. The date provided by these sources are uneven: somereports provide information that others ignore. The tables in this articleare limited to those years for which data are provided.The history of census taking among the European population beganwhen the Pioneer Column reached Zimbabwe in 1891. The aim of thecensus was a population accounting of the number of Europeans whohad survived the ordeal of settling in a new colony. This first census waseasy because after a year of being in the new colony, settlers lived inclose proximity with each other so census taking could be easily doneThe first census of the Pioneer Column showed that there were 1 5ooEuropeans in the column and some unspecified number of their helpersR. MARINDO 147who had travelled from South Africa to the new colony. After this Initialcensus, subsequent censuses were carried out after every five years until1961.The early censuses (those for the years before 1921) were mostlypopulation figures with little mortality or fertility information. The focuswas simply on population growth of the Europeans and there was notmuch interest in the Indigenous population.From the onset, European census covered the European, Asiatic andColoured populations in a de facto household enumeration. These censuseswere carried out by the Department of Statistics which later became the
Central Statistical Office. African population figures were estimated bythe Ministry of Native Affairs based on the number of taxable native men.Village headmen (Sabhukus) provided estimates of the number of menover the age of 15 (taxable age) to the Native Commissioner's Office on ayearly basis. The number of African women and children were made fromthese village estimates by assuming that for every taxable man, therewere three or four women and children (Director of Native Affairs, 1935).Such estimates were likely to be underestimates because they excludedthe elderly, and sometimes village headmen had their own reasons forproviding incorrect estimates.From the onset, different departments were allocated to deal withpopulations of the two races. This separation of data collection isimportant because it shows that data collection for the two races wasbased on two different philosophical foundations which affected thequality and quantity of data collected.Table 1 shows that between 1901 and 1969, a total of 13 Europeancensuses were carried out covering the European, Asian and Coloured
populations of Southern Rhodesia. During this same period, only twoAfrican censuses were carried out in 1962 and 1969. Before 1962, allAfrican population figures were based on estimates.The practical implications of the different philosophical foundationsof collecting data for different races are reflected in the paucity of Africancensus data. The absence of African censuses before 1962 Is clearlystartling when compared to 13 European censuses through the same timeperiod. It appears as If getting reliable figures of the African populationthrough proper household interviews was not a colonial priority before1962. This omission in data collection limits any mortality analysis of theAfrican population before 1962. The absence of African census before1962 also brings into question the issue of how the colonial governmentmade native policies and resource allocation on the basis of broad
estimates rather than proper census figures. In addition to the collectionof reliable population figures of the European population, data on mortalitywere also augmented through the use of a vital registration system.148HISTORICAL ADULT MORTALITY IN RHODESIA (ZIMBABWE)Table 1POPULATION TOTALS AND SUCCESSIVE CENSUS DATES, 1901 TO 1961Census date31st May 190117th September 190429th September 19077th May 19113rd May 19214th May 19265th May 19315th May 19366th May 19417th May 19468th May 19518th May 195626th September 1961May 1962May 1969European1 Populationenumerated1103212 59614 00723 60633 6203917449 91055 40868 95482 386135 596177 124221 504ŠŠAfrican populationestimated500 000590 000670 000740 000860 000930 0001 080 0001 260 0001 400 0001 770 0002 170 0002 990 0003 550 0003 616 600 (1st Africancensus)4 480 000Source: Central Statistical Office (1964), European Life table and cause of death dataTable 2 shows that in addition to broad population figures by year,death rates by disease by year were collected for each year from theEuropean population in the years 1930-1937.Data on causes of death were collected through the vital registrationsystem. In addition, in the remote European farms, vital registrationofficers carried out home visits to collect data (Census Report, 1942)This emphasizes the attention to detail that was applied in collectingEuropean data. The system of data collection therefore differed verygrossly between the African population and the European population. Interms of the epidemiological transition, the data indicate that the majorcauses of death were mostly degenerative diseases. Cancer and heartdisease accounted for most of the deaths in the European populationHowever, the epidemiological transition was not totally towardsdegenerative diseases because pneumonia and malaria also accountedfor a substantial amount of deaths.1 Although the term European census is used in the census reports, it should he emphasl-that the European census included European. Asiatic and Coloured populations Tk.largest proportion of the European. Asiatic and Coloured population w;is clearly tkEuropeans. ' ttl(!R. MARINDO 149Table 2SPECIFIC DEATH RATES PER 100,000 OF 10 MOST COMMON CAUSES OFDEATH AMONG THE EUROPEAN POPULATION, 1930-37SpecificcauseInfluenzaDysenteryTB of therespiratorysystemMalariaCancerDisease ofthe heartPneumonia/bronchitisDiarrhoea193023173129108818333Violent deathsincludingsuicides911931201226484697782810619323382949968474278419331525216571695421731934382619717211962178719355613307285128702410119361645909314692783193744454694134771481Source: Central Statistical Office (1938), The 1937 Census of Kuropeans, Asians and Coloureds.European causes of death.The mortality conditions of Rhodesian Europeans did not differ muchfrom that of other Europeans in colonies. Table 3 shows internationalcomparisons in life expectancy between Rhodesia and Canada, SouthAfrica, US and England and Wales. In the next sections we review some ofthe major concerns that appear to be the push behind the census datacollection in Rhodesia.Table 3INTERNATIONAL COMPARISONS OF LIFE EXPECTANCIESExpectation Southern Canada England & SA USSex of life Rhodesia 1961 Wales 1961 1961 1961MalesFemalese°e20e°e2066.949.47456.268.3551.5174.1756.656850.67455.964.5748.6670.0853.5166.849.773.455.7Source: Central Statistical Office (1964), European life Table No. 2 and Analysis of Cause ofDeath.150 HISTORICAL ADULT MORTALITY IN RHODESIA (ZIMBABWE)CONCERN WITH EUROPEAN POPULATION GROWTH RATE AND ISSUESOF RACIAL PURITYOne outstanding motivation for the collection of population data was theneed to establish the level of population growth of the Europeanpopulation in comparison to that of the African population. All censusesfrom 1901 monitored European population growth through births, deathsand migrations. However, the need for more growth of the Europeanpopulation in comparison with the native population became clear In the1921 census. Increased interest In the European population was probablydue to losses from the First World War, loss due to influenza mortalityafter the 1918-1919 epidemic and the eugenics movement that had takenroot in Europe in the early 20th Century.The perceived need for quality stock2 of people influenced the focuson European growth of the population as opposed to the native populationgrowth.A special population report given by the Director of Population to thelegislative assembly highlighted that the European population growthrate had increased from 3.06% per year in 1911 to 3.74% per year by 1921and this was considered to be a very significant change.The Director of Census In Southern Rhodesia in 1922 put thissuccinctly by stating thatThe popular idea that the native races of Africa are increasing fasterthan Europeans has been long known to be a delusion. It was very fullydiscussed by Prof. J. M. P. Muirhead in a paper called 'Some populationfigures' which he read before the South African Society for theAdvancement of Science in 1906. The proof of this point for Rhodesiamay be given in figures simpler than calculations of per cent differenceswhich are sometimes suspect. In 1911 there were 23, 606 Europeansand 626, 259 indigenous natives. In 1921 the figures were 33, 620Europeans and 761, 790 indigenous natives. In other words there werein Southern Rhodesia at the first date 29,45 indigenous natives forevery European while ten years later the proportion was reduced to22,63 to one (CSO, 1922, 2).The use of ratios is particularly revealing and signifies the extent towhich effort was made to confirm, even to those not technically mindedthe importance of the growth of the European population. By 1925concerns for the growth of the European population had shifted to2 Critical at this time were the work* of Sir Francis Galton (1909). Essays in EugeniaEugenics Education Society); F. C. S. Schiller (1932) Social Decay and Eugenics MoLJ,and the Nazi's theories about racial purity In German. The fact that the eugenics moveZI:was affecting the colonies Is reflected In the Brazilian policy of whitening th*[Thomas Slddmore (1993), Black and While] and Mclaren's (1990) work on Canada eL"!2'Our own master race.R. MARINDO 151concerns about race and maintaining the purity of the British citizenship(CSO, 1927). In the 1926 census, the Director reported that informationon the non-native population was collected by birth place of parents inorder to identify those who were of pure3 British descent. The reportstated that Hebrews had to be excluded from the census because theyfailed to provide the name of the land in which their parents were born(CSO, 1927, part 2, Chapter IX).Concern for racial purity led to an increased focus on the Coloured(mixed race) which had not been obvious in the previous censuses. Inthe 1926 census, a two-paged section was allocated to the discussion ofthe Coloured population. The roots of this concern are clearly eugenic innature. In a parliamentary report of the results of the 1926 census, theDirector of Census made the following remarks:The factor of miscegenation creeps in. This evil exists in our midst andthere are increasing numbers of children who are offspring of whitemen with native women and of Indian men with native women. I aminformed by one of the leaders of the Indian community that the Indiansare building up their own family life in Southern Africa by bringingmore of their own wives. There seems reason to hope that the lattervery undesirable mixture of nationalities is steadily decreasing. It isalso hoped that the weight of adverse public opinion will extinguish thedegrading cohabitation of white men with black women and that whatMrs Sara Millin in her book on South Africa terms 'casual begetting ofthe day' will quickly disappear in Southern Rhodesia (CSO, 1927, 10).The Director of Census then goes on to suggest more rigorous datacollection procedures for selecting out the Coloured population and itsorigin. He suggested that each household questionnaire for the Colouredpeople should have the following extra questions;If both parents are Coloured write 1If one parent is White write 2If one parent is Asian write 3He argued thatBy adopting this or some other similar interrogation, more preciseinformation will be procurable regarding the origin of the colouredpersons, which if carried on from census to census will be ofconsiderable value in dealing with this grave social problem (CSO,i927,section XV, part 3).Other racial aspects of the census are reflected by the fact that onlyhalf a page of the whole census report Is allocated to natives employed3 The purity was based on one being a product of two British parents because the issuewas not only one of whiteness but of citizenship.152HISTORICAL ADULT MORTALITY IN RHODESIA (ZIMBABWE)onuu European farms and mines and nothing at all on African women andchildren while three pages of the report are full of statistics on livestockof the European population which include chickens, turkeys, ducks, horses,cattle and other domestic animals, their numbers and illnesses. Althoughthere was a rinderpest epidemic which destroyed domestic animals in1895 (Southern Rhodesia Official Yearbook, 1924-26) this focus on domesticanimals in 1926 does not seem to be for veterinary reasons.One of the few pieces of mortality data which can be used forcomparing mortality across the two races is the data on mortality amonghospital admissions. Hospitals were separated by race and the Europeanadmissions data came from European hospitals while the African datacame from the native hospitals, so there was a racial divide in thecollection of these data. There are differences in mortality among admittedpatients (Table 4). These differences could be due to the fact that Africanhospitals had poor quality of services compared to European hospitals.In addition, many Africans distrusted the hospitals and tended to gothere only when they were in a critical condition from which they wereunlikely to survive.Table 5 shows the causes of death among hospital admitted patients.It is clear that Africans have higher case fatality rates for all diseasesexcept tumors. This suggests that either conditions for African hospitalswere so poor that patients died of treatable diseases like pneumonia orAfricans were admitted under very critical conditions or both. What thedata also suggest is that Africans were experiencing a differentepidemlological transition to that of Europeans.In terms of epidemiological transition, the European populationexhibits a predominance of degenerative diseases while the Africanpopulation shows higher mortality from infectious diseases liketuberculosis, respiratory infections and digestive diseases.CONCERN WITH AFRICAN MALES AS SOURCES OF LABOURFrom earliest censuses, the colonial government was concerned withharnessing native men as sources of labour for the White households,farms, plantations and mines. A number of reasons account for thisinterest in native labour. British colonies were expected to be economicallyself-sufficient. This meant that in addition to recuperating some of thegains from the colonies to the mother country, the colonies were alsoexpected to take care of themselves by any means possible (Blake, 1977).So to ensure production and profit, the colonies resorted to exploitingthe labour of African males.R. MARINDO153Table 4HOSPITAL ADMISSIONS AND DEATHS 1906-1928, BY RACEYear19061907190819091910191119121913191419151916191719181919192019211922192319241925192619271928CasesEuropeansadmitted Deaths1626156115812 3612 1212 1682 2832 2332 1282 4032 0732 3842 8052 8683 1233 2732 9763 3102 8333 5243 8274 2314 995726986104937710098968876119159104182108122117109143131151219Deathper 1004.424.425.444.414.383.554.384.394.513.663.674.995.673.634.103.304.103.533.854.063.423.574.40Native AfricansCasesDeathsadmitted Deaths per 1009661292161716482 4372 1942 6892 8022 09519262 2042 5182 3642 3302 94931963 0533 0153 9914 30151685 178613715023428528742132238437028325629431631523231240638631946240865363276015.5218.1117.6317.4217.2814.6814.2813.2013.5113.2913.3412.5513.3210.0010.5812.7012.6110.5811.589.4912.6412.2012.39Source: Department of Statistics (1938), The Official Annual Social and Economic Conditionsof the Colony of Southern Rhodesia.In order to force native men to work for the European population,Rhodes in his Glen Grey Act of 1898 in South Africa had come up with theidea of introducing a special labour tax on all native men (Smuts, 1930).This was later changed to a hut tax in Rhodesia which was paid by alladult native men as their contribution to 'good government' (RhodesianNative Affairs Report, 1932). Smuts justifies the use of native labour forthe development of the colony and for the good of the native by sayingBy temperament the natives have not much initiative and if left tothemselves and their own tribal routine they do not respond very wellto the stimulus for progress. They are naturally happy-go lucky and areTable 5RELATIVE MORTALITY BY TYPE OF DISEASE FOR AFRICAN AND EUROPEAN DEATHS AMONG ADMISSIONSTuberculosisTumorsNervous systemCirculatoryPneumonia andother respiratoryDigestiveGentto-urlnaryPuerperal stateSkin andcellular tissuesBonesCongenitalmalformationsEarly InfancyExternal causesHi-definedTotal19233.949.604.6210.007.293.792.191.64ŠŠŠŠ0.90Š3.541S244.3810.495.708.4510.923.003.021.92Š2.78ŠŠ1.81Š3.85Europeans19253.2612.996.0615.2110.704.032.214.560.601.33ŠŠ2.32Š4.0619263.5418.053.6920.2911.060.921.934.84Š1.82ŠŠ1.180.353.4219274.1213.643.5917.476.492.922.54ŠŠŠŠŠ2.040.773.57192314.8510.7310.3315.0525.7019.3912.505.000.67ŠŠŠ3.652.9010.58192415.608.3416.7416.6626.7314.278.646.251.533.13ŠŠ3.620.8911.58Africans192510.3914.4113.9714.8119.1012.098.4929.180.70Š33.33Š4.524.899.49192611.9211.1612.7530.0029.51133510.533.842.474.32ŠŠ4.131.1012.64192715.1210.0015.1033.7226.2010.5813.146.581.792.50ŠŠ4.811.5612.20Z3)3"NiCD*. Th« U»ni»*ht« Colonial Sectatary 00301. Official Y^arhtx** of //i«- Colony of Southern Rhodmni,R. MARINDO 155not oppressed by the stirrings of that divine discontent which havemade Europeans the most unhappy but most progressive of all humans.They are easily satisfied and a very little goes a long way with them. Asworkers they are slow, unintelligent and essentially imitative. Theyhave little foresight and display little forethought. But thesecharacteristics make them take readily to a routine which is settled forthem by the white employer. For thousands of years they have beenaccustomed to the domination of their chiefs and therefore they readilyaccept the firm handling, the lead and the mastery of the white employer.It fits in with their character and their age-long training (Smuts, 1930,30).But more importantly, the employment of natives was seen as acivilizing tool and Smuts goes on to state thatIt follows that the easiest most natural and obvious way to civilize thenative is to give him decent white employment. White employment ishis best school; the gospel of labour is the most salutary gospel for him(Smuts, 1930, 31).Details of mortality and morbidity were collected only from nativemen working in mines (Table 6). It was also stated that the aim was to getas much information as possible about the natives in employment inorder to ensure that they were adequately taxed and to ensure that theirsickness did not affect productivity (CSO, 1926, 639).From the onset, the native population was not considered in a holisticmanner but was separated into labour and non-labour, useful and not-useful. Mortality and morbidity information was collected only from the'useful' population not from the whole population. The colonialistapproach to dealing with natives using divide and rule was thenentrenched into population data collection from this early date. Onegroup of men from which mortality data were collected were Africanminers. Data were collected on mortality rate due to mine accidentsunderground and deaths due to disease. Table 6 shows these mortalityrates for 1906 to 1929.Data on African miners' diseases and deaths were collected for allsingle years until 1953. What is clearly outstanding is the amount ofefficiency with which these data were collected for each year especiallywhen compared with the absence of information on mortality or causesof death of African women and children.Information on mortality of miners was also collected by country oforigin of the miners (Table 7) again highlighting that where there wassufficient interest in the African population by the government, mortalitydata could be collected.Table 6MORTALITY AMONG AFRICANS EMPLOYED IN SOUTHERN RHODESIA MINES FROM 1906 TO 1933.YearAverage no.of workersNumber of deathsdue to diseaseYearlydeath rateNumber of deathsdue to accidentsYearly accidentdeath rateTotaldeathsTotal yearlydeath rate19061907190819091910191119121913191419151916191719181919192019211922192319241925192619271928192917 38126 09830 86532 72137 82637 90934 49435 54336 10037 92840 52038 46132 76630 29637 66937 60535 71837 48241 28639 3864161741 63542 94046 SI 21 1631486139713831682108510737838978329117003 62950759968961850466550559859575687565.9156.9445.2642.2744.4728.6431.1123.4924.8521.9422.4818.01110.7616.7315.9018.3219.0713.4416.1112.8214.3514.2917.6119.691571021321611821641631581351591721498890759486105891059194941109.033.914.284.924.814.334.734.713.744.194.243.832.692.971.992.502.402.802.162.672.182.312.12.35111111111332058852954486424923694603299108384171759767478376760975461068968985098575.9460.8549.5447.1949.2832.9535.8328.2028.5926.1326.7321.85113.4419.7117.9020.8221.4716.2518.2615.4916.5316.5519.8021.0430ooz33OmCO03mSoorce: Department of Statistic* (1947). Statistical Yearbook of the Colony of Southern Rhodesia.R MARINDO 157Table 7MORTALITY RATE AMONG NATIVE EMPLOYEES BY COUNTRY OFORIGIN FROM 1930-37Country of origin 1930 1931 1932 1933 1934 1935 1936 1937Southern Rhodesia 7.13 7.22 7.26 6.39 5.84 7.84 5.0 5.76Portuguese East Africa 17.09 15.03 10.32 10.95 11.11 7.7110.10 10.80Northern Rhodesia 17.18 11.13 8.04 7.44 10.46 13.56 10.2110.73Nyasaland 18.55 16.67 11.71 12.96 12.20 12.58 12.74 10.62OtherTotal22.0115.157.4512.619.499.543.839.2459.15.356.6311.801.499.441.429.26Source: Department of Statistics (1938), The Official Annual Social and Economic Conditionsof the Colony of Southern Rhodesia, Ministry of Finance and Commerce.In examining the data on cause of death it is impressive that so muchinformation was collected per year. What is striking, however, is theselective nature of these data when compared with the absence of anydata from Africans in the villages. This further suggests that it was notimpossible to collect data, but decisions were made which justifiedcollecting data from select parts of the population which were deemed tobe useful for the colonial government.CONCERN WITH EPIDEMICS AND OTHER INFECTIOUS DISEASESConcerns relating to epidemics and infectious diseases also differed forthe different races. In the earlier censuses before 1920, no efforts weremade to collect data on European deaths and births as these were obtainedthrough vital registration. The 1921 census however, asked questions ondeath in order to capture the losses in the European population due tothe influenza epidemic of 1918 and how it had affected mortality rate.Table 8 shows the mean death rates collected from the 192I.Europeancensus.Death rates increased for all the ages in 1918 but were more thandouble for the 15-44-year age groups. This suggests that the epidemicwas an adult epidemic rather than a child epidemic. Although the agepattern of mortality for influenza years appears rather atypical and couldsuggest omission of childhood deaths, there is no supporting evidencefor this since almost all European deaths were registered. It is impressivethat at this very early date, less than 30 years since its establishment, thecolony was so organized that they managed to collect good mortalitydata for the Whites. No data exists on mortality due to influenza of the158HISTORICAL ADULT MORTALITY IN RHODESIA (ZIMBABWE)Table 8DEATH RATES FOR THE EUROPEAN POPULATION FOR NON-UEA ,NFLUENZA AND INFLUENZA YEARS4AgegroupMean Annualdeath 1918deathrate (10-yr raleaverage) 1000Percentincrease ofdeath ratein 1918Actual number ofdeathsAveragefor 9 yrs 1918Under 11-45-1415-2425-3435-4445-5455-6491.913.63.65.48.411.716.724.3100.122.65.812.325.622.423.628.88.9266.1861.11127.77204.7691.4541.3218.5259.8930.3315.3318.4441.4456.6643.1123.00756132541601278035Source- Central Statistical Office (1927), The 1926 European. Asian and Coloured PopulationCensus'of Southern Rhodesia, European deaths data.native population although Patterson (1980) suggested that about 25,000Africans died in Southern Rhodesia during the epidemic.5In addition to the influenza epidemic, two diseases, syphilis andleprosy seemed to have been of major concern to the mining companiesspecifically but also in the general population. Illife (1987) in his book,The African Poor, provides some convincing evidence of the debilitatingeffects of leprosy in some of the African colonies.Although syphilis was recorded In the Medical Director's Report as amajor cause of death among miners in 1924, the mining cause of deathrecords do not support this (Table 9) but instead indicate that pneumoniawas the major cause of death and syphilis did not rank among the fivemajor causes of death. However, this does not discount it as a devastatingcause of morbidity. However, because morbidity data among minerswere not collected, it is difficult to substantiate this.4 Report of the Director of Census of Population for the .Southern Rhodesia Colony, ig^l,6 Patterson (1980) states that 'The influenza epidemic of 191R-1919 was certainly the singlegreatest short term demographic catastrophe In the continent's history. Nothing else, £,,slavery, colonial conquest, smallpox, cercbrosplnal meningitis, the rinderpest paniooMrof 1880 and 1890 nor the great trypanosomlasis after 1900 killed so many Africans In «short a time'(p. 404). *R. MARINDO 159Table 9CAUSE SPECIFIC MORTALITY RATE AMONG NATIVE MINE WORKERSBY DISEASE, 1930-1937DiseaseMalariaScurvySyphilisPneumoniaTB and otherdiseases of thechestDiarrhoeaHeartDebilityInfluenzaOther diseasesTotal1930.68.04.298.321.21.70.40.13.952.4315.151931.54.17.285.971.39.51.45.48.682.1412.611932.28.11.424.691.19.61.39.06.281.519.541933.39.27.293.991.18.72.56.12.231.499.241934.58.08.324.22.94.67.53.13.581.389.351935.46.29.295.69.91.73.54.171.251.4711.801936.46,13.324.83.84.59.48.14.251.409.441937.36.06.264.75.70.53.41.18.451.569.26Source: Department of Statistics (1938), op. cil.INTERNATIONAL AND INTERNAL MIGRATIONMigration played a very critical part in the population dynamics ofRhodesia. Not only were the settlers migrants, but they used populationmovements to control and design the colony as they saw fit throughresettlement as they took over fertile land for their own use. Migrationpolicies therefore differed by race.International migration of female Europeans and Asians wassncouraged in order to control the mixing of races. The eugenics roots of:his approach were discussed in a previous section. European immigrationwas encouraged and land was offered for free to those Europeans who:ame to settle in the colony. The success of the colony was judged by thelumber of Europeans who came to the colony.Rural to urban, migration of the native population was strictly:ontrolled. Male natives were allowed to migrate to towns and mines tovork in order to pay taxes to the government. Rural to rural migrationvas not allowed without the permission of the Native Commissioner.Native female migration was controlled because they were supposed toinsure the continued existence of genuine African village life. Smutstates that he agreed with Rhodes's basic idea that160 HISTORICAL ADULT MORTALITY IN RHODESIA (ZIMBABWE)I have emphasized the importance of native institutions, of keepingintact as far as possible the native system of organization and socialdiscipline. As long as the native family home is not with the white menbut in his own area, so long the native organization will not be materiallyaffected . . . And it Is the migration of the native family, of femalesand children to the farms and towns which should be prevented. Assoon as the migration is permitted the process commences which endsin the urbanized detribalized native and the disappearance of nativeorganization (pp. 99-100).The different policies with regard to migration of native men andwomen led to shortages of women in the mining and plantation areas.This led to the development of prostitution at the mines and plantations.Van Onselen (1976) argued that not only was the government aware ofprostitution in the mine compounds, they encouraged it. He states thatThe mass of workers came to mines without their women, and from thevery earliest days of the industry, prostitutes were a feature ofcompound life. Since their services did much to attract and stabilizelabour, mine management and state were unwilling to eliminate it (VanOnselen, 1976, 49).Turshen (1987) argues that resettlement affected the diseaseenvironment by breaking the boundaries between the local populationand disease vectors. The resettlement of local populations from highlandsas the land was being appropriated for the European population, meantthat the native populations were sometimes settled in areas which wereendemic for malaria and tsetse flies and from which the natives haddeliberately kept away.6 This must have had an impact on the morbidityand mortality of the native population.In addition, the forced migration of men into towns leaving women Inthe villages created labour shortages in the African villages. Most of themen were paid single person's wages which meant that women not onlyhad to produce food for consumption by the children and old people Inthe villages but had also to supply supplementary food for the men inurban areas or mine compounds. This is likely to have increased femalemorbidity and mortality. Heywood and Thornton (1987) and Turshen(1987) suggest that the emigration of males also led to an Increase Infertility as women needed child labour in the villages to replace themissing men. Frequent pregnancies and the extra load could have furtherIncreased female mortality.6 Meredith Turshen gives a very convincing argument of how resettlement of localpopulations forced them to clear bushes'and break down barriers between vectors andhuman settlement areas. She uses colonial Tanzania as a case study.R. MARINDO 161Writing on colonial Tanzania, Turshe (1987, 190-91) stated:increased work loads, poorer nutrition, more pregnancies (the newmode of production also increased demands for child labour providingcouples with incentives for larger families), and the declining socialand political power of women contributed to rising mortality amongwomen in the African villages.African males who migrated to mines, plantations and urban areaswere exposed to new diseases. Because of overcrowding and poor housing,many became exposed to infectious diseases especially respiratorydiseases which were not common in the villages. The shortages of womenalso meant more exposure to prostitution and associated sexuallytransmitted diseases which had not been common in their villages.Patterson (1980) suggests that the movement of European troopsinto Africa also led to the spread of diseases for example, both small poxand the 1918 influenza epidemic could be traced back to seaports thatwere busy with war time traffic: Boston in Massachusetts; Brest in Franceand Freetown, Sierra Leone. So colonial policies and the resultant migrationpatterns in a number of ways interacted in affecting the mortality andmorbidity of African men, women and children.Migration for labour reasons of native men from other parts of Africato come and work in the mines was also encouraged. These miners wereusually paid less, many suffered disease at the mines because of badconditions. Table 7 shows the mortality by country of origin of theminers who were brought to work in Rhodesian mines.POLITICAL UNREST AND THE IMPETUS FOR COMPREHENSIVE AFRICANCENSUSAs many African men were forced into urban areas looking for work, thetowns became meeting places for political organization through the mixingof men from different ethnic groups.Hodgkin, writing about the 1950s towns of Africa, suggested thatThe point has been made that the new towns of Africa act as solvents,weakening traditional social ties and loosening the hold of traditionalbeliefs and values. This is partly true: but the positive function oftowns is no less important. By providing opportunities for a greaterdegree of specialization, towns enable men to acquire new skills andpowers. By mixing men from a variety of social backgrounds, theymake possible the discovery of new points of contact and interest.Around these interests there develops a network of associations throughwhich for the first time men come to think of their problems as socialrather than personal: as capable of solution by human action ratherthan part of a natural order (Hodgkin, 1956, 5G).162 HISTORICAL ADULT MORTALITY IN RHODESIA (ZIMBABWE)The effect of exposure, education and media meant that urban Africansbegan to have more contact with Africans from different countries. Theattainment of political independence in some of the early colonies likeGhana and Nigeria were not isolated events but interdependent politicaloccurrences influencing the rest of Africa. Colonial boundaries failed tocontrol these influences.Political organization was also made easy by the apartheid system inwhich Rhodesian towns had been developed as racial towns with Blacktownships and White suburbs. The existence of a Black and White townmeant that African towns housed two nations, not only in Disraeli'ssense, a nation of the rich (or relatively rich) and a nation of the poor butalso in a special colonial sense; the nation of White and nation of Blackand the correlation of being Black and being poor and being White andbeing rich was enough to stimulate a spirit of African radicalism whichtended to identify the claims of the poor against the rich with claims ofthe Blacks against the Whites.In response to these urban political disturbances, it became necessaryfor the government in power to have some indications of the scale of theproblem by having some estimate of the numbers of native populationsboth in towns and in the rural areas. A demographic sample survey ofAfricans was then carried out in 1953-55. For a government with such ahistory of detailed census taking, this African demographic survey wasoutstanding for its poor quality and incorrect information. It was statedby the Director of Census in 1959 in his report of the African DemographicSurvey to the legislative assembly thatIn designing the survey, the first consideration that had to be borne inmind was the limited professional staff available for training the fieldofficers. Secondly, thought had to be given to the fact that large scalepopulation movements were taking place in many districts as a resultof various resettlement schemes. Under these circumstances it wasconsidered advisable to stagger the work over three years (CSO, 1959,3).Although the focus of the survey was on Africans, it was stated thatthe data were bad because 'many of the Africans were very reluctant toprovide information about their social lives' (p. 54). So very rudimentarydata were collected on fertility and mortality and indirect methods wereused to make estimates. The samples chosen were mostly from areasthat were easily accessible by car leaving out some of the remotepopulations. The final report is poor, there are many typing errors andthe figures do not add up. The report tells very little about t he demographyof Africans.Special demographic surveys for Africans in urban areas were carriedout. The results of these urban surveys were also very poor, because inR. MARINDO 163anticipation of negative repercussions of being found in towns, largenumbers of African females and children migrated back to the rural areasduring this period and moved back immediately after the survey, distortingthe findings of these town surveys. Those who remained in towns hidfrom interviewers in an effort to curtail their presence in the towns.COLONIAL LEGACIES: IMPLICATIONS FOR CURRENT MORTALITY DATAZimbabwe attained political independence in 1980. The taking over of thesettler government by the Black government meant in some casesinheriting government bureaucracies with the same inbuilt inertia. TheCentral Statistical Office, which had been in charge of collecting data onthe White population began to collect data on the whole population usingthe same procedures that they had established previously. The first post-independence census began by doing away with the concept of race inpopulation statistics and thereby made it impossible for continuity ofdata analysis between the pre-independence and the post-independencecensuses. In practical terms, many of the areas that had been excludedfrom previous colonial censuses because of lack of accessibility werealso excluded from the first post-independence census in 1982. The firstpost-sindependence census was never fully analysed. Only 10 percent ofthe data was analysed and the rest were corrupted during storage. Whatthis reflects may be lack of expertise in the administrative storage of databecause most of the African officers involved in data collection in thepost-independence era had previous experience but most of them in avery junior capacity as data collectors rather than data storage.In terms of the methods of data collection, there has been no visibleattempt to come up with different ways of collecting data from populationsthat are not numerate. Although many people who are not formallyeducated may have problems with counting, there are traditional ways ofcounting which can be refined and applied in a form of participatory datagathering. Both census officials and demographers appear to be stuck inan mode where indirect estimation appear to be the only way forward. Indata analysis, census reports rely on applying population models manyof which do not fit African data.The racial differences in quality between European data and Africandata are similar to current differences in data quality between rural andurban areas. More care is taken in collecting urban data, where all theadministrative powers are located. In addition, vital registration officesare urban based, another legacy from colonialism which continues withoutany effort to establish village based vital registration systems.From the villagers' side, the distrust of census officers leads to a lotof lying and attempts to cheat the system. The distrust of vital registration164 HISTORICAL ADULT MORTALITY IN RHODESIA (ZIMBABWE)and perception that it is not useful procedure for Africans date back tothe colonial government's divide and rule strategy Š the 'them and us'attitude.7 Many rural populations have never been adequately educatedto use census data for their particular areas which could assist them toplan their businesses, location of schools and even vegetable markets. Tomany villagers, census activity is something planned by government inthe city and then imposed on them. They do not see the census as a toolthat they can use for their own planning. It is something that is donebecause the government says so.These instrumental uses of demography which have some roots inthe past, must be understood so we can begin to see them more clearly inthe present.CONCLUSIONFrom the historical data, mortality data differ between the races both inquantity and in quality. Different offices were allocated to data collectionŠ with the Central Statistical Office collecting mortality data from theEuropean population while the Ministry of Native Affairs estimated Africanpopulation figures. Different methods of collecting data were used, in theEuropean case de facto household censuses were carried out periodicallywhile African population figures were estimated from the number ofnative men who were aged 15 years and above (taxable age).These administrative differences had implications for the quality,quantity and usefulness of the data collected. In terms of Europeanmortality, it is possible to trace a mortality history from 1901 to 1961.There is a dearth of African mortality data, with virtually no data onAfrican women and children mortality. This paucity of data creates a gapin the mortality history of the African population.A deliberate selection process was used in collecting data from Africanmales employed in mines. These data, although useful by themselves interms of employment related mortality, tell us little about mortality ofAfrican men in general.Both the quantity and quality of data was affected by deliberateselection (of European population versus African population, minersversus other men), deliberate omission (of African women, children andthe elderly in the villages) and of over emphasis (European population).All these factors indicate that data were distorted to give a certain viewof reality.7 Marindo-Ranganai and Zaba (1996) showed how colonial legacies affected the conservationof wild animals in Northern Zimbabwe. In evaluating the Communal Areas ManagementProgramme for Indigenous Resources (CAMPFIRF.) programme in Chapoto Ward, theyfound that villagers considered animal conservation as a 'them' and not 'us' activity Šalegacy that came from colonial times.R. MARINDO 165To understand historical mortality, we also have to acknowledge theglobal influences that were taking place at that time. The eugenicmovements in Europe influenced racial politics in Southern Africa. Whatmakes the Rhodesian situation and the South African situations moreoutstanding is the gross expression of racial discrimination through theapartheid system. It is a clear expression of uncontrolled state power. Itis important to realise this because colonialism can not be perceived assomething out of the global context.This fact is important for African demography, because it suggeststhe need for critical evaluation of motives of international and nationalorganizations, many of which may not be grossly racist but may bedriven by motives that come from certain 'traditional' notions of others'social and demographic behaviour. There is need for a critical analysis ofmotives that drive demographic data collection. This caution isparticularly pertinent because of the expansion of the field of demographyinto intimate behavioural studies for example, in the areas of HIV andfamily planning where racist and ethnic notions can be easily justifiedunder 'humanistic' or 'public health concerns'.What the study shows in the final analysis is that what is at stake isnot only the 'colonization' of demographic statistics but also the'colonization' of those who collect the data.The importance of understanding the social and political contextwithin which demographic data are collected cannot be understated. Tounderstand contemporary mortality in Zimbabwe, one has to acknowledgethe contribution of colonial history whose legacies still persist in presentday data collection.Demographic change, or in this case mortality change cannot beexplained by one factor, to do that would be reductionist.Kwame Anthony Appiah in his book called In My Father's Housesuccinctly discusses this issue when he states:I would want to resist the reduction of this work to a single theme . . .and in addressing that situation I would not want to bury the manystories in a single narrative. This claim has become a postmodernistmannerism: but it strikes me as, in fact, also a very old and sane pieceof wisdom (p. xi, Preface).Colonialism itself does not explain the higher mortality of Africans,yet the colonial administration's treatment of African males as labour,the created poverty in the rural areas and increased working burdens ofwomen through the capitalistic appropriation of African male labourcombined with new disease environments and the poor hospital facilitiesfor the African population all contributed to higher mortality amongAfrican men, women and children. What is clear however, is that racialdiscrimination acted as a thread that perpetuated and tied these factors166 HISTORICAL ADULT MORTALITY IN RHODESIA (ZIMBABWE)together and influenced the historical picture of adult mortality inRhodesia.Finally, censuses are not necessarily objective administrative tools.To analyse them in an ahistorical, apolitical approach is to provide abiased view. Censuses are political tools that have both an outcomecomponent and yet can also define and redefine social categories.BibliographyGovernment Documents/PublicationsCentral Statistical Office (1962) The 1961 Census of European, Asian, andColored Population of Southern Rhodesia (Southern Rhodesia,Salisbury).Central Statistical Office (1954) The 1951 European, Asian and ColoredPopulation Census of Southern Rhodesia (Southern Rhodesia,Salisbury).Central Statistical Office (1964) European Life Table No. 2 and Analysis ofDeath By Cause (Southern Rhodesia, Salisbury).Honorable Colonial Secretary (1930) Official Yearbook of the Colony ofSouthern Rhodesia, No. 2 1924-1928 (Salisbury).Honorable Colonial Secretary (1932) Official Yearbook of the Colony ofSouthern Rhodesia, No. 3 1921-1928 (Salisbury).Department of Statistics (1938) The Official Annual of Social and EconomicConditions of the Colony of Southern Rhodesia, Ministry of Finance andCommerce (Salisbury).Department of Statistics (1947) Statistical Yearbook of Southern Rhodesia1942 (Salisbury).Department of Statistics (1945) Statistical Yearbook of Southern Rhodesia(Salisbury).Central Statistical Office (1959) Demographic and Sample Survey of theIndigenous African Population of Southern Rhodesia. 195.3-55 (Sails buiy,Rhodesia).Central Statistical Office (1927) The 1926 European. Asiatic and ColoredPopulation Census of Southern Rhodesia (Salisbury. Rhodesia).Central Statistical Office (1989) Zimbabwe. Demographic and Health Survey1988 [CSO Harare and DHS (Institute of Resource Development/MacroSystems, Maryland, USA)].Zimbabwe National Family Planning Council (l'Wti) History of FamilyPlanning in Zimbabwe (Harare, ZNFPC Publications).Central Statistical Office (1984) The Zimbabwe Reproductive Health Survey(Harare, Westinghouse).Southern Rhodesia (1974) Migration Report (Salisbury. GovernmentPrinters).Ministry of Health (1986) Zimbabwe Medical. Dental and Allied Professions/k? (Harare, Government Printer).R. MARINDO 167Ministry of Health (1986) Zimbabwe, Health for All Action Plan (Harare,Causeway).Books and ArticlesAppiah, K. A. (1992) In My Father's House: Africa in the Philosophy ofCulture (New York, Oxford University Press).Beach, D. N. (1980) The Shona and Zimbabwe 900-1950 (New York, AfricanaPublishing Company).Blake, R. (1977) A History of Rhodesia (London, Eyre Methuen).Bruwer, A. J. (1965) Zimbabwe: Rhodesia's Ancient Greatness(Johannesburg, Keartland).Cohen A. (1959) British Policy in Changing Africa (Northwestern UniversityPress, USA).Galton, F. (1909) Essays in Eugenics (London, The Eugenics EducationSociety).Garlake, P. S. (1970) Great Zimbabwe (London, Thames and Hudson).Š (1983) 'Prehistory and ideology in Zimbabwe', in J. D. Y. Peel and T. 0.Ranger (eds.) Past and Present in Zimbabwe (Manchester, ManchesterUniversity Press).Gayre, R. (1972) The Origin of Zimbabwe Civilization (Salisbury, GalaxiePress).Smuts J. C. (1930) 'Livingstone and after', in African and Some WorldProblems Including the Rhodes Memorial Lecturer Delivered inMichaelmas Term (Oxford, Clarendon Press).Š (1930) 'Native policy in Africa', in African and Some World ProblemsIncluding the Rhodes Memorial Lecturer Delivered in Michaelmas Term(Oxford, Clarendon Press), 70-103.Gervais, R. (1987) 'Creating hunger: Labor and agricultural policies inSouthern Mosi, 1919-1940', in Dennis Cordell and Joel Gregory (eds.)African Populations and Capitalism: Historical Perspectives (TheUniversity of Wisconsin Press).Greenhaugh, S. (1996) 'The social construction of population science: Anintellectual, institutional and political history of twentieth centurydemography', in Comparative Studies in Society and History, 38, (i), 26-66.Godwin P. (1996) Mukiwa: A White Boy in Africa (Harare, Baobab).Heligman, L. Nancy Cohen and Ozer Babakol (1993) 'Shifts in the structureof population and deaths in less developed regions', in James N.Gribble and Samuel H. Preston (eds.) The Epidemiotogical Transition:Policy and Planning Implications for Developing Countries (WashingtonD. C, National Academy Press).Heywood, L. and John K. Thornton (1987) 'Demography, production andlabor: Central Angola, 1890-1950', in Dennis D. Cordell and Joel W.Gregory (eds.) African Population and Capitalism: Historical Perspectives(London, Westview Press/Boulder).168 HISTORICAL ADULT MORTALITY IN RHODESIA (ZIMBABWE)Hodgkin, T. (1956) Nationalism in Colonial Africa (London, FrederickMuller Ltd.).Wife, J. (1987) The African Poor: A History (New York, Cambridge UniversityPress).Maclver, D. Randall (1906a) 'The Rhodesian ruins', in Geographical Journal,27.Manning, P. (1987) 'Local versus regional impact of slave exports onAfrica', in Dennis D. Cordell and Joel W. Gregory (eds.) AfricanPopulation and Capitalism: Historical Perspectives (London, WestviewPress/Boulder).Marindo-Ranganai, R. and Zaba, B. (1996) 'Animal Conservation and HumanSurvival: A Case Study of the Tembomvura People of Zambezi Valley,Northern Zimbabwe' (Harare, Centre for Applied Social Sciences,University of Zimbabwe).Mclaren, A. (1990) Our own Master Race: Eugenics in Canada (Ontario,McClelland and Stewart Inc.).Mugwagwa, N. 0. (1989) 'Evolution and growth of family planning inZimbabwe', in Proceedings of the Colloquium on the Impact of FamilyPlanning Programs in Sub-Saharan Africa: Current Issues and Prospects(Legon, University of Ghana).Painter T. M. (1987) 'Making migrants: Zarma peasants in Niger, 1900-1920', in Dennis D. Cordell and Joel W. Gregory (eds.) AfricanPopulation and Capitalism: Historical Perspectives (Madison, TheUniversity of Wisconsin Press).Patterson, D. K. (1980) 'The demographic impact of the 1918-19 influenzapandemic in Sub-Saharan Africa: A preliminary assessment', in AfricanHistorical Demography Vol, II, Proceedings of a Seminar Held in theCentre of African Studies (University of Edinburgh).Raeburn, M. (1978) Black Fire! Accounts of the Guerrilla War in Rhodesia(London, Julian Friedmann Publishers).Turshen, M., (1987) 'Population growth and the deterioration of health:Mainland Tanzania, 1920-1960', in Dennis D. Cordell and Joel WGregory (eds.) African Population and Capitalism: Historical Perspective's(London and Boulder/Westview Press).Van Onselen C, (1976) Chibanr. African Mine Lalx>ur in Southern Rhodesia1900-1933 (London, Pluto Press Limited).