Employment Generation in Africa: How Have Females Shared? This research wasfacilitated by'a grantfrom the University of Illinois Research Board. Graduate student research assIStants were Rene Kinimoz and Tom Gillard-Byers. Jean M. Due. Professor of Agricultural Economics University of fllinois at Urbana-Champaign Utaflli Vol VI/I No.1, 1986" Journal of the Faculty of Arts and Social ScienCes University of Dar es Salaam: Introduction Although it is known that women produce 60 per cent of the food in Afri~ can(l), not much is known about the importance of women in the labour force (even in the agricultural labour force) or to what extent females have benefited in employment growth since independence. Several studies have been under- taken to ascertain the causes of labour force (LF) migration from rural to ur~ ban area in individual countries (2-13); some of these studies have included both male and female migration, others have been primarily concerned with ~ migrl¥ion. Recent International Labour Office (ILO) and World Bank data now allow a comparison of employment growth from 1950 to 1978 by coUntry and by sex. How have women shared? This paper looks .at these African data to ascertain the rates at which employment was increasing for the continent as a whole and the distribution of employment by sex. The ILO data brought together population and labour Foree (LF) partici- pation for each country, using standard concepts through cut. LF is defined as "all employed and unemployed persons, i!1clu4ing those seeking work for the first time. Iteovers employers, pe.rsons working on their own ac- count, salaried employees, wage earners, unpaid' family workers, members of producer's co-operatives and members of the armed forces. It covers all persons workjng regardless of age" (14, p.vii). The World Bank' definition of the LF includes only ~rsons from 15 t06S years of age, whereas ILO includes anyone working regardless of age .. -' When one looks at the proportion of the population in the LF by sex using the ILO data with the standard concepts, the variationoi females in the LF by country is astounding. In Algeria, only 1.9 per cent of-the total f~e popu- lationwas considered to be in the LF in 1910 compared to 51.7 percent in Mali. For Africa as a whole, 25.0 percent of the total female population and 52.7 per cent of the male population were in the LF in 1970~ (14. Tables 1 and 2). Obviously, these ILO data depend on the categoriettOS defmitions utilised in each country; in most countries 'homemaking and child care' are not consi- dered "work"! However, other factors also influence female participation out- side the bl)usehold. One of these is religion, with the Moslem religion being less supportive of women working outside tbe household than most others; another is tradition and culture; for wage employment, level of edUC9tion is an important entree. Both the World Bank and ILO divide the LF into three broad categories. agriculture. indUifry, and service.2 Table 1: The percentage of the LF in each of these categories for African oller time is shown by the unbracketed figut:es; unbracketed figures show th~ percentage of females in each category. 27 Table 1. Summary for Africa of Distribution of Total Labour Force by sector and Per-. centage of Women in each Sector (Brackets)* Agriculture Industry Services flo % 0/0 Africa 19502 81 (34) 7 (15) 12 (33) 1%02 77 (35) 9 (IS) 14 (31) 19702 72 (35) 11(18) 17 (34) 1~ 79 8 13 197~ 69 13 18 .* Source: 2Source: 14, Table 3. 3Source: 16, Table 19 and 21, Table 35. Agriculture . As expected .the JX:rcentageof the LF in agriculture declines as the level ot employment riSes; in the 29-year period tllepercentage in Africa has declined from. in to 69 per cent. It apPears to'r:each 50 per cent at annual per_~apita 'incomes' 'of around $900 in 1978 (figure 1). Again the percentage of the LF in agriculture'varied widely from 91 per cent in Rwanc,laand Niger to 30 per cent in South Africa and 20 percent in Libya in 1979I(Tabie ~). Although the World Bank uses a different definition of the LF, their distribution of tIle LF by category in 1960 is remarkably similar to the ILO data.lmd. the.more recent year (1979)allows the use ora more recent time period. It should be noted that the World Bank data are for 45 African countries3 while the ILO data are fO! 49 . . In most countries of Mrlca, persons in the agricultural LF are self-employed rather than being wage employed or tenant farmers. The percentage of females in the agricultural LF varied widely from less than 2 percent in Libya to 56 per cent in Botswana in 1970(Thble2). Note that the percentage of females in the total agricultural LF in Mrica l¥ls reIJ?aine~stable at 34 to 35 percent from 1950 to 1970;these percentages are shown In the bracketed figures in Thble 2. (These bracketed figUres indicate the percentage of women in each of the categories; for (:ltaIDplea third of the LF in agriculture and services but only 18 per cent of that in industry was female in 1970). '!\va countries in which religious differences do not appear significant, and which are geographically side by side, have very different percentages of women .in the agricultural-LF, these are Benin with 14 per cent and Togo with 37 per cent in 19701Such differences are hard to explain except as outlined at the top of page 2. The average annual growth rate of the total African LF was 1.85 percent from 1950to 1970; the l1UInberof men in the LF has grown slightly faster than the number of women (1.87 compared with 1.81) for the same period, ILO projections are that the rate of total LF participation will accelerate each de- cade until 2000, reaching an average annual rate of 2.7 percent in the 1990's; however, ILO projections are that males will enter the total LF at a faster ari~ nual rate than females for each decade. 28 ; -to , ••• _1- • ,.~ 6 __ ! _~ _ . ._,----- " . ~~ ' .-- T ••••• - ~ _ ~._._-- ..,. .." +' .> U .... " ::> ev '0 -- -, - --- -- ------ - ~ '>, '" -- .0 CJ t- '" o: '- "'> -lev 0 .. +, ~ -l .r; Ul ----0 C' : & '" .,.::" :z: "'C '" . 0: .QJ~ ... V> 0' .... V> "'OJ 0 h +'~ -c: ev'~ UCO .... t+- • t~ !... t.:' i s "'" ~" ev+' .- ... -_." . "- ",. ev >- u OJ !... ~.~ > ev .... -'-ev • ..•.. "-'" --s .;:; '. .... ::>VI u 0' !... ;:;t t:' <: ~ ~ 8 ..,o .. o Cl 29 Table 2. Distribution of the Total Labour Force bi. Sector and Percentage of WODlell ill Eacb Sector (sbown in brackets), Africa, 1960, 1970,1 and 1978*2 eoitntry Agriculture Industry .Services OJ, OJ, OJ, Africa 196()2 77 (35) 9 (15) 12 '(33)' 19702 72 (35) 11 (18) 14. (31) 19793 69 ( ) 13 ( ) 18 ( ) Algeria 1960 67 (1) 12 (4) 21 (lO) 1970 61 (2) 15 (4) 24 (9) 1979- 30 ( ) 25 ( ) 45 ( ) Angola 1960 69 (5) 12 (2) 19 (21) 1970 64 (6) 14 (3) 22 (21) 1979 60 ( ) 16 ( ) 24( ) Benin 1960 54 (17) 9 (51) 37 (85) 1970 50 (14) 12 (49) 38 (84) 1979 46 ( ). 16 ( ) 38 ( ) Botswana 1960 92 (50) 3 (7) 5 (4) 1970 86 (56) 4 (9) 10 (36) 1979 83 ( ) 5 ( ) 12 ( ) Burundi 1960 90 (48) 3 (7) 7 (22) 1970 87 (48) 4 (9) 9 (25) 1979 84 ( ) 5 ( ) 11 ( ) Cameroun 1960 88 (52) 5 (3) 7 (12) 1970 8S (48) 6 (4) 9 (14) 1979 ~~I 7 ( ) 10 ( ) Cape Verde 1960 68 (5) 10 ( ) 22 (14) 1970 62 (6) 12 ( ) 26 (15) 1978 () . ( ) ( ) Central 1960 94 (53) 2 (11) 4 (16) African 1970 91 (53\ 3 (23) 6 (18) Republic 1979 88 .( 4( ) 8 ( ) Chad 1960 95 (24) 2 (3) 3 (15) 1970 90 (25). 4 (3) 6 (9) 1979 85 ( ) 7( ) 8 ( ) Congo 1960 52 (41) 17 (23) 31 (42) (I9?O~ 42 (36) . 21 (27) 37 (43) 1979 3S ( ) 26 ( ) 39 ( ) Egypt 1960 ~ (6) 12 (3) 30 (12) 1970 SOl(6) 19 (5) 27 (15) 1978 51 ( ) 26 ( ) 23 ( ) Ethiopia 1960 8S (39) 5 (36) 7 (26) 1970 ~ (36) 6 (37) 10 (28) 1979 . aO ( ) 7 ( ) 13 ( ) 30 Gabon 1%0 85 (45) 7 (Ii) 8 (~) 1970 81 (44) 9 (5) 10 (27) 1979 79 ( ) Wi ) 11 ( ) Gambia 1960 8~ (48) 7 (14) ~ (iln 1970 82 (49) 8 (20) 10 (21) 1979 19( ) 1O( ) 11 ( ) Ghana 1960 64 (4~) 14 (13) 22 (52) 1970 58 (39) 17 (38) 25 (51) 1979 54 ( ) 20 ( ) 26 ( ) Guinea 1960 88 (51) 6 (21) 6 (1O) 1970 85 (44) 8 (26) 7 (11) 1979 82 ( \, 11 ( ) 7 ( ) Guinea 1960 90 (3) 4 ( 6 ( ) Bissau 1970 87 (4) ~( J P { \ 1979 93 ( ) 1( ) 6 ( ) Guinea 1960 84 (3) 6 ( 10 (12) Equat9xil!! 1970 80 (3) fj ( ) 12 (pr, 1978 ( ) ( ) ( ) Ivory Coast 1960 88 (51) 2 (5) 10 (42) 1970 84 (45) 3 (5) 13 (46) 1979 79 ( ) 4( ) 17( ) Kenya 1960 86 (38) 5 (9) 9 (21) 1970 82 (37) 7 (10) 11 (22) 1979 '78,( ) JO( ) 12 ( ) Lesotho 1960 93 (48) 2 (18) 5 (36) 1970 90 (46) 3 (16) 7 (36) 1979 87 ( ) 4 ( ) 9 ( ) Liberia 1960 80 (39) 10 (4) 10 (18) 1970 75 (38) 12 (6) 13 (22) 1979 71( ) 14 ( ) 15 ( ) Libya 1960 53 (2) 16 (13) 31 (6) 1970 32 (2) 22 (9) 46 (5) 1978 21 ( ) 27 ( ) 52 ( ) Madagascar 1960 93 (48) 3 (15) 5 (32) 1970 89 (48) 4 (17) 7 (34) 1979 87 ( ) 4 ( ) 9 ( ) Malawi 1960 92 (40) 3 (11) 5 (22) 1970 89 (40) 4 (11) 7 (25) 1979 86 ( ) 5 ( ) 9 ( ) Mali 1960 94 (42) 3 (70) 3 (14) 1970 91 (48) 4 (72) 5 (16) 1979 88 ( ) 5 ( ) 7 ( ) 31 Mauritania 1960 91 (4) 3 ( 6( ) 1970 88 (5) 4 ( 8 (3) 1979 85 ( ) 5 ( ) 1O( ) Mauriti.lls 1960 40 (20) 26 (6) 34 (27) 1970 34 (22) 25 (8) 41 (25) 1979 30 ( ) 24 ( ) 46( ) Morocco 1960 62 (3) 14 (25) 24 (19) 1970 57 (10) 17 (19) 26 (20) 1978 53 ( ) 20 ( ) 27 ( ) Mozambiq- ue 1960 81 (35) 8 (5) 11 (6) 1970 73 (33) 13 (3) 14 (8) 1979 67 ( ) 17 ( ) 16 ( ) Namibia 1960 62 (22) 17 (3) 21 (42) 1970 56 (24) 21 (2) 23 (40) 1978 ( ) () ( ) Niger 1960 95 (9) 1( ) 4 (9) 1970 93 (10) 2( ) 5(8) 1979 91 ( ) 3 ( ) 6( ) Nigeria 1960 71 (43) 10 (24) 19 (48) 1970 62 (38) 14 (31) 24 (51) 1979 55 ( ) 18 ( ~ 27 ( ) Reunion 1960 46 (7) 21 (5) 33 (45) 1970 38 (7) 21 (8) 41 (45) 1978 ( ) ( ) () Rwanda 1960 95 (SO) 1 (6) 4 (26) 1970 93 (50) 2 (6) 5 (30) 1979 91 ( ) 2 ( ) 7{) Senegal 1960 84 (44) 5 (11) 11 (18) 1970 80 (44) 6 (16) 14 (23) 1979 76 ( ) 10 ( ) 14 ( ) Sierra Leone 1960 77 (41) 12 (7) 11 (29) 1970 71 (40) 15 (10) 14 (35) 1979 66 ( ) 19 ( ) 15 ( ) Somalia 1960 88 (32) 4 (5) 8 (18) 1970 85 (32) 6 (6) 9 (20) 1979 84 ( ) 8 ( ) 8 ( ) South Africa 1960 32 (12) 30 (7) 38 (45) 1970 31 (33) 30 (12) 39 (49) l'978 30 ( ) 29 ( ) 41 ( ) 32 Sudan 1960 86(1) 6(12) 8(11) 1970 82(1) 8(13) 10(11) . 1979 7B( ) 10 ( ) 12( ) Swaziland 1960 89 (04") 04(17) 7 (27) 1970 81 (049) 6 (17) 13 (040) 1979 '2 ( ) 9( ) 39 ( ) Tlnzania 1960 89 (042) 4 (10) 7 (".2) 1970 1979 86 (39) 83 ( , S (12) 6( ) 9 (2') III ) Togo 1960 80 (37) 8 (.la) 12 (66) 1970 7J (37) II (34) 16 (70) 1979 68 ( ) IS ( ) 17 ( ) Tunisia 1960 '6 (1) 18 (IS) 26 (10) 1970 50 (I) 21 ~17) 29 (II) 1978 45 ( ) 24( ) 31 ( ) Uganda 1960 39 (36) 4 (10) 7 (2S) 1970 86 (36) S (10) 9 (27) 1979 83 ( ) 6 ( ) 11 ( ) Upper Volta 1960 92 (47) .s (71) 3 (IS) 1970 87 (46) 8 (72) 5 (17) 1979 83 ( ) 12 ( ) .s ( ) Zaire 1960 83 (51) 9 (-~) 8 (7) 1970 79 (52) 11 (6) 10 (9) 1979 75 ( ) i3 ( ) 12 ( ) Zambia 1960 79 (34) 7 (14) 14 (040) 1970 73 (31) 9 (20) 18 (45) 1979 68 ( ) 11 ( ) 21 ( ) Zimbabwe 1960 69 (32) 11 (13) :0 (32) 1970 61 (32) 13 (14) 26 (33) 1979 :so ( ) IS ( ) 25 ( ) • Sources: 2. ILO 14, Table 3. 3. World Bank 16, Tables 19 arid 21, Table 35. 33 . 'Table 3: Average Annual Growth Rates (Percent), of Males and Females in the Labour Force in Africa by Sector, Selected Years* Total LF Agricultural Industry Services Total M F Total :M F Total M F Total M. F 1950-19601 1.6 1.7 1.4 1.1 1.1 1.2 4.5 4.4 4.9 4.2 4.7 3.2 1069-19701 2.1 2.0 2.2 1.5 1.5 1.5 5.7 5.1 8.9 4.6 4.0 5.9 1960-19702 1.8 1970-19802 1.9 1970-19s03 2.2 2.3 2.1 1980-1m3 2.4 2.5 2.3 1990-z0003 2.7 2.7 2.7 .*Sources: 1. ILO, 14 Tables 1 and 4. 2. World Bank, 16, Table 19. 3 .. ILO Projections, 14, Table 4. The agricultural LF had grown at a fastl:f annual rate for the decade of the 1960's than a decade earlier as has indus1!t:ies and services. Females were' entering the total LF and industries and services at a faster annual rate than males in the 1960's. Although there are no: projections by sector for the post-1970 periods, ILO predicts that males' entry into the LF will be at a faster-- rate than females for those periods. The countries where ILO predicts that fe- males will enter the total LF at a faster percentage rate than males are the D!edOminaI]tIy Moslem cQuntries.4Thus the au~hor challeng~ th.e ILO projec- tion regarding rate of male to female entry; given changes in attitudes, levels! of education and opportunities, and the low levels of femafe participation in the Moslem countries currently, I predict that women will continue to enter the LF at annual percentage rates at least equIvalent to men. I further argue that the continued liberation of Moslem women will strengthen my estimate. Un- fortunately, there are no recent World Bank estimates to compare with the ILO projections nor are there ILO estimates by sector. However, it is known that if employment opportunities do not open in industries and services, agriculture will have to absorb the surpluses. Employment in industries and services was growing at a much faster annual rate than agriculture for both decades. Women's Participation It is noted from the bracketed figures in Table 1 that African women form a higher percentage of the LF in agriculture and services (where self-employed is more important) than in industry {where wage employment is the norm,. Ap- proximately one-third of the LF in agriculture and services was female and this has been constant throughout the 1950-1970 period. In contra"t, only 15 per- cent in 1960 and 18 peroI2) + .3839x(; - .0234x7 + .0123xg - .0734?<9 - .11~10 (1.6623) (-2.5602) (2.0770) (- .2967) (-1.6449) 'R2 = .72 38 These variables explained 72 percent of the variation. in Y (R2 = 72) (the per- centage of the LF in industry and services), but only si:/(.variables were signifi- cant at the 90"10 confidence level. This equation can be interpreted as stating that, as GNP per capita increases by.$l per annum, Y increases by .07 percent; as the annual percentage growth in population increases by 1 percent, Y decreased 9.13 peI:Cent, as energy consum:etion per capital increases, Y decreases .02 percent, and as net inflow of loans' Ih<:reased by $1 million, Y increases, .01 percent. As the index of per capita food production increases one percen- tage point, Y decreases .32 percent, and as the percentage of the Moslem popu- lation increases 1 percent, Y decreases .11 percent. The negative sign on per capita food production was surprising. The hypothesis was that, as one increased food production per capita, more of the LF would be released for industry and services, just the opposite is suggested by the estimated equation. Perhaps the change in the index of per capita food prodnr:tiolJ. is a proxy for relative incomes in agriculture and industry; as the index improved, a larger percentage of the LF stays in agriculture. Seventy-six percent of the variation in Y was obtained by using all of the ind~pendent variabl~s above, except energy consumption per capital (X7) and net mflow of loa~s (XJ and substituting the percentage of students of secon- dary school age enrolled (XII)' The equation now became: (2) Y = 39.6431 + .0319x1 - 5.7086x2 + .0458x3 - .1927x