*Labour Force Participation and Unemployment in Lobatse by Nesim Tumkaya Introduction Following Indepel1(lencein 1966, Botswana has embarked on a series of National Developnent Plans aimed at allocating her limited resources efficiently.1 The preparation and irrplarentation of these plans, however, have been harrpered by a lack of or insufficiency of derrographic, econanic and social data, particularly those on employment and unE!lployment. So serious the shortage of data on E!lployment and unemploymentis that these indices are not fully knownrut estimated indirectly by the National Developnent Plan,2 even tit\!»19hpolicies at creating employment are accorded a central role in the Plan.3 Partly to shed scxre light on the issues of employment and unE!lploymentand partly to train students in field research procedures, the present lal:our force survey was conducted in IDbatse in May 1981. It is based on a randan cluster sample of 291 households c::cruprising1491 persons (6.8 percent of the estimated population).4 As a partial analysis of the survey, we examine below the levels and patterns of laJ::our force participation in IDbatse for males and females, detennine the magnitude and age pattern of unemployment, and study the characteristics of the unemployed persons. Concepts of Employmentand UnE!lployment According to the conventional def ini tion, the E!lployed and unemployed 1r.I:Jrkers c::cruprisethe "civilian" seCJll¥i!llt of the laJ::our force,5 \'.hich includes all persons \'.ho furnish the supply of laJ::our for the prOduction of econanic goods and services during the reference period. The E!lployed persons are those v.ho were 1r.I:Jrkingduring the reference period, including unpaid family rrembers, or v.ho had a job from \'.hich they were temporarily absent. The un- employed are those \'.ho were not 1r.I:Jrkingduring the reference period ru t \'.ho were seeking WJrk for payor profit, including those \'.ho never WJrked before ru t v.ho were seeking 1r.I:Jrk.Also included arrong the unemployed are persons \'.ho were not seeking 1r.I:Jrk during the reference period, because of temporary illness, because they had made arrangements to start a new job, or because they were on ternp:Jrary or indefinite lay-off without pay. 6 \'ihere employment opportuni ties are very limited, the unE!lployed also includes persons v.howere not 1r.I:Jrkingand not looking for WJrk because they believed that no jobs were open. It.rhe author wishes to thank the United Nations, Department of Technical Co- operation for Developnent and the University of Botswana for their support in carrying out this research, and the students of Field Research in Derrographycourse v.hose dedication and diligence made the survey possible. Any shortcomings on this paper, however, rests entirely on the author. ai CIl CIl <11 <11 ~ 11 l;:1~ <11 ....., ..... O>Or"'lOLf"l 0 ......... ~ III ~ ~ <11 ,;; &; Lf"lOr"'lOLf"l Lf)L.OMl.t"lOl;;Jl I I '" ~ Lf"l<1l N13 >Ol-l 1<11 :5~ I~ 1i '0 ia'O ~ 11 ~ <11 ONr---\Oo~ OOr-:MONI Lf"lN ~~ r"'l ... ~ ~~~ <11 ..... CO"l::1'~OO::>O • I r"'l 0 I1l l-l 01::l 01 k •..-1 k Lf"l NCOo>O ... 0 "'''' 0> I"- I"- .~~ .jJ~ .g l-l .e11i <11 U' III ui I-< l-l l-l g- <11 ~~ . ".: l-l <110 13 • ~ CIl ~, j qoCO\O"=2'r-Qr--. MO'\I..ON~ ~ N° Lf"l 1-<13 8:5~ !!d ~ ~ .~01~ '@ l-l CIl <11 13 k ~ III W ~ ..... ::l ...-1 .Sl CIl 01 ~ '0'0 U <11 CO~CO"'OO>'" 0> k k .jJ .;,:l-l .jJ ~ ~ MCOr---Lf)M~(\",) r"'l ~~o ID l-lO .....III CIl 0:; 0 r"'l ~<11~ ~ ~ ..... .--I a :5 ~ <11 .... 01 0 O"l-r-t 8. k ..... .~ >< O>O>NNLf"lOCO ...... lfl 1-<01 III .--I •..-1 ".: ..... .--I .5 •..-1 iE'~ 0:; i _q-N_ till-l 1Il . ..-1 <11 ~ u :J CIl till-l 13 l-l 1-<01 .e i ~i .... 0 <11 k~6 III <11 .......... 13 l-l ~ ~~ O'\r---\ONcYlCOr---. N \,0 ,f9l-lE >-1 ~ <11 0 BSB l-l ~~ .jJ ~ ..... ~ 1~ <11 lflo> ... N",Oo> -oo::r-.:::r- lfl ... M ~v.l <11(1) ~ ~~ ~ ID u tiJ :5.... 1 .~ CIl 0.. CIl (I) 81:5 0 ::I 1 1~ ~ ]13::1 .jJ <11 ,f9.jJ<1I ID .--I O'\r---Nqor---~r---. I"- CIl w til ..... .... Ill U s~~ ~E .8 -\.Or---t.f)N~N ~ l"- N l-l I-< 0> ~ :;:l <11 .~ ~~ •..-1 ij ~% B l-l ~ ~~ ~ 1~ O""'"O'\(X)lflMNM ~ ... co ~<11'@ gj 13 U' k •..-1 01l-l .~ III .~ ~ .~ '0 U. gj ,Q gl-l0,Q......5 .--I'8 @ J 0>0>0>0>0> ,.....Nf"I"') l-l ~ I I I I I III Qj ~ 0 k kill ~~g~~g~ ~ nf'M- ,Q H 0 H.c: J 2 The al:::ovedefini tions have teen cri ticised as unsuited to the social and economic conditions of the developing nations, and alternative concepts were proposed.7 For example, Myrdal rejects the concepts of "unemployment"or "underernployrrent" entirely as inadequate to descri!::e reality, and proposes use of concepts that take into account lal:cur participation, duration of v.ork and lal:::ourefficiency. 8 Similarly, J. Weeks considers the conventional notions of ernployrrent, unernployrrentand underernployrrentas irrelevant to the conditions of the developing countries and recanrrends the utilization of silTpler I::ehavioural concepts that distinguish !::etweenmenbers of the society that v.ork or do not v.ork at all, the periods during the day, week, IlDI1thor year in which v.ork is done, and the in tensi ty and effectiveness of the v.ork.9 Criticisms to the conventional concepts focus on their disregard to the social and cultural factors that in developing countries !::ear a great influence on the meaning of v.ork, type of v.ork and the division of lal:our anong meml::ersof society. Although such criticisms are not entirely baseless, their satisfaction is extremely difficult in an operational definition that can !::eapplied empiri- cally. SOCieties differ from each other in tenus of \\ark canprehension, division of lal:::our, and in the way they I::ehaveto satisfy various activities so that a canprehensive definition that can satisfy the ah:>vecriticisms and still !::eapplicable to all developing societies is perhaps impossible to devise. In the same vain, although the conventional definitions are not perfect, they are presently the IlOst feasible and recarmended ones for use.10 In this study the conventional definitions of employmentand unemploymentare utilized.11 Any bias arising from a possible mismatch !::etweenthe concepts and the realities is minimal !::ecause the study was done in an urban area which also is the largest ernployer in the manufacturing sector of the country. 12 Latour Force Participation Amongthe sample of 482 males and 532 females ten years old and over, 339 males and 250 females v.ere in the lal:xJur force; ¥ielding crude activity rates of 50.7 and 34.4 percent respectively (Table 1). 3 Since very young persons cannot participate in economic activities, a refined acti vi ty measure is calculated by taking the population of ten years and al:::oveonly. 14 The resultinq "refined" activity rates arrount to 70.3 and 47.0 percent for males and females respectively, clearly indicating a greater propensity aIrong males to join the lal:::ourforce than females. The propensity to join the lal:ou.r force, hov.ever, is highly dependent on age for !:oth sexes. Indeed, as the age-specific lal:our force participation rates show, the greatest propensity to join the lal:::ourforce is in ages 40-49 for males and 30-39 for females (Table 1). Amongmales, the participation is alrrost universal in all ages !::etv.een20 to 60, l:u t rot so high in ages !::eyond60 and !::elow20. Females' lal:::ourforce participation rates are ITU.lchlov.er than those of males and different also in tenus of a9= pattern. Lal:::o..1r force participation rates are low at ages 10-19 for !:oth sexes, rot they rise sharply in the subsequent ages (Table 1). H~ver, while the rates of males remain at peak levels fran age 25 to 60, those of females drop substantially after age 35 till reaching zero at ages teyond 60 (Figure 1). 3 The al:ove labJur force participation rates resemble the so-called Southern African pattern of lato.rr force participation for rrales and the South African-Namibian pattern for females. 15 The resemblance of the Lobatse lal:n.1r force participation curves to those of the neigh1::orring countries - despite differences in definitions and age groupings - is not coincidental mt is due to the similarities in the darographic, econanic, social and cultural factors that determine the pattern and level of lalxlur force participation. 16 Enployrcent and Unemployrrent Al:ou.t82 percent of the male labJur force in Lobatse was employed during the reference period and 18 percent ~re unemployed. Arrongfemales, the employ- rrent rate was I1U.lch lower (54 percent) and the unemploymentrate rruch higher (46 percent). SUchunemployrrent rates are high by international standards, altl-Dugh not unique. For exarrple, the joint IPale-female unemployment rate was reported as 26.6 precent in urban Algeria in 1966, 30.8 precent in Kaduna city (Nigeria) in 196317 and 21.0 percent in Lagos in 1971. Similar results ~e obtained in Botswana in 1974 Yhere unemployrrent rates of 18 and 48 percent respectively ~e found for IPales and females in three peri- urban areas. 19 A striking feature of unemploymentis that it does not affect everyJ::ody equally. For example, the youngest persons are affected rrost acutely by high unemployrrent, whereas those in adult ages are not so highly affected (see Table 1 and Figure 2). The vulnerability of the youngest persons to unemployrTa1tis not confined to Lobatse tut observed in =y other countries as weU.20 Another striking observation refers to the very low unemployrrent rates arrcng the middle-aged IPales in Lobatse, especially those 30 to 50 years of age. The same cannot be said for females however, for than unemployrcent is high in all ages altlxlugh relatively lower in ages 30 to 40 than in other ages. Characteristics of the Unemployed The observed age and sex differentials in unemployment rates indicate that unenployment affects in different ways the IPales and females and also tlxlse in different ages. By examining the other characteristics of the unemployed ~ IPaygain insight into the causes of unemployrrent. First, ~ ask Yhether the fanale \lneIlIlloyrrent was unduly raised by the in- clusion of tlxlse wh:) ~re not seeking work due to temporary illness or dis- ability and those wh:) tlxlught that jobs ~e not available (there ~re 4 males and 17 feIPales in this status, see footn::>te of Table 1). Reroval of these persons fran the openly unemployed group does not change the unemploy- rrent rates significantly and does not eliminate the male-female differen- tials; the resulting rates equal 17. 3 percent and 42. 1 percen t respecti vel y for males and feIPales. Therefore, ~ may conclude that the rrale-female differentials in unemployrrent rates are genuine regardless of the definition utilized. 4 TABLE 2: PERCEl\'TAGE DISTRIBUTION OF THE JOB-SEEKING UNEMPLOYED l->DRKERS BY MARITAL STAWS AND SEX MARITAL STAWS NEVER WIrovED, DIVORCED SEX MARBlED TOTAL MARRIED OR SEPARATED MALE 77.6 20.7 1.7 100.0 FEMALE 61.2 34.7 4.1 100 TABLE 3: PERCENTAGE DISTRIBUTICN OF THE JOB-SEEKUX; UNEMPLOYED \~RKERS BY RElATIONSHIP TO HEAD OF HOUSEHOLD AND SEX RELATIONSHIP SON/ IN-LAW SEX HEAD SPOUSE PARENT OTHER TOTAL DAUGHTER MALE 5.2 22.4 5.2 31.0 19.0 17.2 100.0 FEMALE 16.3 10.2 30.6 21.4 11. 2 10.3 100.0 5 Madtal status and Relationship Examination of marital status of the job-seeking unerrplOYed21persons in- dicate:>that the majority of them are never married, especially the male (see Table 2). When these persons are further examined fran the standpoint of their relationship to head of household, nost of the males afPear to be parent, spouse or in-law to the head; whereas nost of females are either daughters or parents, and to some extent, heads of oousehold thenselves (see Table 3). This finding has important implications in showing that the majority of the unemployed persons, especially the males, are not heads of h::>useholdthanselves and are therefore not subject to the grave responsibi- lities of providing for families under oonditions of unerrployment. Educational Attainment The unemployed IDrkers appear to have varioos educational backgroonds (Table 4). In general, feroales have ~tter educational attainment than males, and for I:x:lthsexes the majority are prirlary sCOOolleavers, i.e., those who have not a::mpleted Standard 7. Our attention is drawn here to an ~rtant issue: have these primary schJol leavers indeed left school permanently and joined the labour force or are they in the job market on a short-rem basis? Although ~ have no direct information to illuminate this question, it is likely that a great deal of these persons ~e seasonal IDrkers who started looking for jobs during the long schJol vacation when this survey was con- ducted. Consequentl y, ~ may oonclude tha t the lal:x:lur force participation and the unerrployment rates in Lobatse IDUld have ~en lo~r than the present ones had the survey ~en oonducted during the school session. TABLE 4: PERCENl'AGE DISTRIBUTION OF THE JOB-SEEKING UNEMPIDYED w::lRKERS BY EDUCATIONAL ATTAINMENT AND SEX COMPIEI'ED GRADE OF SQiOOLING SEX tQ.lE STANDARD STANDARD FOR1 1 AND NOT 1T06 7 AOOVE STATED 'IUl'AL MALE 31.0 31.8 17.2 13.8 5.2 100.0 FEMALE 21.4 33.7 25.5 17 .3 2.0 100.0 6 FIGURE 1: LABOUR FORCE PARTICIPATION RATES BY AGE AND SEX; LOBATSE 1981 Lobatse,Males Lobatse,Females Labour force Southern African pattern, participation Hales rate (percent) South African-Namibian 100 pattern,Females r-'-'-'-'- I 90 i I i I "'\.\. 80 .I '. I '. 70 i ,-- - - -------'" \. , ,I , , " I f ,, ,, I , I \. ; '. 60 ,i " I " ", \. '. i I , \ ; I ',- i I " 50 i I I " " i / ", i ! I " '~ , I 40 ! i t ..... \ ! ;1' :/ \ \ 30 ! /1 \ I : I \ i h , ;', -.f ', ! ,:, \ I :, \ 20 , 1 h ...i '. \ \, i ! . 10 I \ , .........~\ ".\ '.,, 10 20 30 40 50 60 ~o AGE 7 WOrkExperience AI::out40 percent of the job-seeking males and 45 precent of the females did rot have any previous Y.Orkexperience. As expected, a large part of them belonged in the youngest ages, especially anong females (Table 5). Previous Y.Orkexperience is an asset that usually makes it easier to find subsequent jobs. Therefore, the concentration of inexperienced job- seekers in the young ages strengthens our in'pression al:out their seasonality ani partly explains their high ~loyrnent; rowever, this explanation does not apply to the elderly male unenployed Y.Orkers. Similarly, it does not fully explain the feIl'ale unenployne1t \otlich ranains high in ages \\here the proportion of the inexperienced is relatively low (e.g., 40-49 ani 50-59). TABlE 5: INEXPERIENCED JOB-SEEKERS AS PEOCENT OF 'IDl'AL JOB-SEEKERS BY AGE AND SEX SEX AGE MALE FEMALE 'IDl'AL 10-19 68.4 72.2 70.3 20-29 37.5 44.2 42.4 30-39 50.0 37.0 39.4 50+ 0 25.0 7.1 'l'Ol'AL 37.9 44.9 42.3 Meth::ld of Job Search If sare rrethods of job search are considered to be IlOre effective than others in locating a job, then differentials in rrethods of job search may partially explain the unenployrrent rate differentials between the tY.O sexes. Data 00 the rrethod of job search anong the unenployed show that "asking enployers directly" was the IlOst camon rrethod utilized by l:x>thsexes in Lobatse (Table 6). However, these data have illus tra ti ve value only; they canrot reveal the relative effectiveness of the vaIious rrethods, because the IlOre effective rrethod Y.OUldbe expected to transfer persons from unenployment into enployrrent, therel:1y leaving a small number of persons to quote it. An inquiry on the rrethod of job search arrong the enployed persons W)l.lldbave been IlOre valuable in detennining the relative effectiveness of various rrethods of job search, alth::>ugh such question was not utilized in the present survey. 8 FIGURE 2: UNEMPLOYMENT RATES BY AGE AND SEX; LOBATSE,1981 Unemployment rate (percent of labour force unemployed) 90 80 /0 60 50 -- 40 30 20 10 10 20 30 40 50 60 10 AGE Note: The female unemployment curve was not extended beyond age 60 because the sample did not contain anybody in the labour force in these ages. 9 TABLE 7: PERCENl'JIGE DISTRIBUTIOO OF JOB-SEEKIt-K; UNEMPLOYED mRKERS BY DURATION OF JOB SEAOCH AND SEX SEX DURATION OF MALE FEMALE TOTAL JOB SEl\RCH 1-4 WEEKS 24.1 11.2 16.0 5-8 WEEKs 6.9 6.1 6.4 9+ WEEKS 69.0 82.7 77.6 TOTAL 100.0 100.0 100.0 TABLE 8: PERCENT1IGE DISTRIBt1I'ION OF '!HE EXPERIEN::ED AND INEXPERIEOCED UNEMPLOYEDmRKERS BY DURATION OF JOB SEAOCH AND SEX MALE FEMALE DURATION OF JOB SEl\RCH EXPERIENcED INEXPERIENCED EXPERIENCED INEXPERIEOCED 1-4-WEEKS 19.4 31.8 13.0 9.0 5-8 WEEKS 5.6 9.0 7.5 4.5 9+ WEEKS 75.0 59.0 79.6 86.4 TOTAL 100.0 100.0 100.0 100.0 10 TABLE 6: PEOCENTAGE DISTRUBITION OF THE JOB-SEEKING UN:E1>1PLOYED WORKERS BY METHOD OF JOB SEAlUI, AGE AND SEX I MALE FEMALE ASKING ASKING AGE OTHER TOTAL OTHER TOTAL EMPLOYERS EMPLOYERS 10-29 88.6 11. 4 100.0 75.4 24.6 100.0 30-49 83.3 16.7 100.0 77.8 22.2 100.0 50+ 100.0 0 100.0 75.0 25.0 100.0 IDr STATED 85.7 14.3 100.0 66.7 33.3 100.0 TOTAL 89.7 10.3 100.0 75.5 24.5 100.0 Duration of Job Search The job-seeking unemployErlpersons ~re askErl for how long they have been looking for \-.Ork. The results shownin Table 7 indicate that the job searcl' is a lengthy process; about 69 percent of the unemployed males and 83 per- cent of the females spend nine weeks or longer in job search. Significant differentials exist, ~wr, between males and females with respect to wr duration of job search and the previous \-.Orkexperience. InexperiencErl females spend rrore t.ine on the job search o::mparErlto the experienced ones, v.hereas arrong males the opposite situation is true. Males wittout \-.Orkex- perience appear to have spent less tiIre on the job search than the ex- periencErl ones (Table 8). This omervation strengthens oor earlier impression that some of the male jOb seekers are new entrants to the labour force searching for seasonal errployrnent during the long school vacation. Conclusion This research aiIred primarily at exploring the lal:x:>urforce participation and unemployment levels and patterns in Lobatse in May 1981, and at describ- ing the characteristics of the unemployed persons. The data was analysed by means of bivariate and !lUlti-variate contingency tables. Since the principal purpose was exploratory, tests of significance and rrore definitive measures of relationship ~e rot deered necessary at this stage, although there is clearly a need for rilprous statistical analysis of the deter- minants of labour force participation and unerrployment for lxlth males and females in Lobatse and in Botswana. This study has revealerl that the labour force participation patterns for Lobatse are similar to those found in the neightouring countries. The male pattern is characterizerl by high and continuous participation betv.een ages II 20 and 60 and is said to resemble the SOUthernAfrican pattern of lal;our force participation. The female pattern characterizes lower levels of participation that pealeei1 ages 25 and 35 and drop thereafter, resembling the SO-called SOUthAfrican-Namibian pattern of fanale lal;our force participation. The survey also yielded very high unerrployrnentrates for both sexes, albeit Imlch higher rates for females than males. The ooenployed males are found to be concentrated in the very young and very old ages, leaving relatively low unercployrrentrates in the interIrediate ages. A large segment of the 00- errployed males are non-heads of households, unmarried, an::1 primary school leavers who are in the lal;our market for the first t:i.rne. This leads us to conclude that the adverse econanic, social an::1 psycological effects of 00- enployment are sarev.tlat mitigated in Lobatse because nost of the unerployed are not heads of households who usually bear the chief responsibility of caring for the family. There are also indications that sore of the young unerployed persons are transient in the job market looking for enployment during the schcol vacation. The female uneuployrrent in Lobatse is serious, hov-A2ver, not only because it is higher than that of modes rot because also it affects all ages. Like males, a great number of the unerrployed females are not heads of households, unmarried, and withou t previous jOb experience. Both males and fanales spend relatively long periods in job search. The high unerrployrrent rates in urban areas of the develop~g countries have often teen attriroted to excessive rural-urban migration. Urban areas, it is argued, create an image of present or future opportunity that attracts rural populations< The knowledge of high unenployment rates in cities, the tough conpeti tion for jobs, an::1 the extended periods of job search do not deter the prospective migrants \'.he reckon that their efforts at locating jobs in the city will eventually succeed, leading thereby to a belief that the "expected" returns to migration are larger than those to staying in rur al areas. 23 A series of erpirical studies in Africa seem to support this hypothesis. o.rr Lobatse data, lacking migration information, do not allow a similar test, however, we may speculate that rural to urban migration in Lobatse is certainly an :inp::lrtant factor in increasing the size of the labour force as well as the ooerrployed. Weare also unable to describe the characteristics of the rural-urban migrants with these data and the ways in which they accarodate themselves in the city. However, jUdging from the relatively large proportions of "in-laws" and "others" in the relationship to head of household character- istics (Table 3), we may speculate that manypersons, especially males, care from rural areas to Lobatse an::1 stay with friends or relatives \\hile searching for jobs. If present errployment oppJrtunities are to continue in the future, the chances of securing jobs in due time appear to be gOOdfor males rot not so for females. 12 FOOINOTES 1. Botswana had four National Development Plans up to 1979. The fifth covers the period 1979-85. 2. The lal:our force is estimated as the total nurrber of persons in ages 15 to 64; the unenployed are assurred to conprise the residue between the lal:our force and the sum of migrant It.Orken>abroad, the fornally enployed persons and an equal nurrber assumed to be their dependants . See, Ministry of Finance and Developrreti Planning, National Develofrnerlt Plan 1979-85. Garorone: Cbvernrrerit Printer, 1980, pp. 17-20. 3. Ibid., pp. xi, 43. 4. With a de-facto population of 11,936, Lol::atse was the fifth largest settlerrent in Botswana in 1971. Its survey time population was estimated as 21,970 by projecting the 1971 population into 1981 accord- ing to the average growth rate, of 6.28 precent per annumthat prevailed during 1964-71 5. The "armed forces" are not included in the present analysis. 6. United Nations, Principles and Reo:mrnendationsfor Population and Housing Censuses Statistical Papers, Series M, NO. 67. Newyork: United Nations, 1980, pp. 93-4. 7. An excellent sunrnary of these criticisms and the alternatives suggested is given in Guy Standing, Larour FGn::eParticipation and Develof!llerlt. Geneva: 110, 1978, pp. 25-54. 8. Myrdal, Gunnar, Asian DramaVol. 2. Hamondswrth, Middlesex: Penguin Books, 1968, pp. 961-1027. 9. Weeks, J., "[X)esenployment matter?" in Jolly, R., et al., Third ~rld Errployrrent: Problems and Strategy. Hamondswrth, Middlesex: Penquin Books, 1973, pp. 61-65. 10. United Nations, Principles and Recomrendations, pp. 93-94. 11. Errployed rrembers of the larour force VJerethose who have wrked anytime "last VJeek"and those who have not wrked but who had a job from which they VJeretenporarily absent. The unenployed VJerethose who VJere either lookina for wrk last week or who VJerenot looking for wrk, but wanted to wrk, because they believed that no jobs VJereopen or VJere tenporarily sick or disabled. 12. National Develo~nt Plan, pp. 203-4 13. Patio of the labour force to the total population in all ages. 14. United Nations, Meth:Jdsof Analysing Census Data on ECO~c Acti~ties of the Population Population Studies, No. 43. NewYork: Unlted Nations, 1968, p. 9 13 15. Altlulgh this female pattern primarily characterizes the urban areas of SOuth Africa and Namibia, it is observOO also in many other areas, in- cluding Europe and North AnErica. See Denti, Ettore, "Sex-age patterns of la1:our force participation by urban and rural p:>PJlations." in International Lal:our Review \bl. 98, No.6. Dec. 1968, pp. 525-50. 16. Factors affecting la1:our force participation are discussed in United Nations, Determinants and COnsequences of PoPJlation Trends Vol. 1, PoPJlation Studies, No. 50. Newyork: United Nations, 1973, W. 299- 310. 17. 'lbdaro, M., "Incorre expectations, rural-urban migratioo and employment in Africa", in Enploynent in Africa. Geneva:.!.!£, 1977, p. 46. 18. Different studies oowever give different rates, ranging fran 7 percent to 25 precent. See Fapohunda, 0., Ellployrrent and Unalployrrent in Lagos. The Hague: Institute of SOcial Studies, 1977, W. 8-9. 19. According to that study the unenployrred were defined as "tOOse persons \04"0 have no personal ineate from alploynent or self-errployrrent, and Yklo say that they are actively seeking w::>rk". Ibwever, when defined to include only tOOse \04"0 have looked for w::>rkduring the previous 30 days, the unerrployrrent rates drop to 15 and 35 percent respectively. See, Central Statistics Office and Ministry of Finance and Developnent Planning, A Social and Econcmic Survey in Three Peri -Urban Areas in Botswana, 1974. Gal:orone: Q:>vernnent Printer, 1974, W. 42-3. 20. See, WJrld Bank, Latour Fbrce, Ellploynent and Latour Markets in the Course of Eoonanic Developnent i'«:>rkingPaper No. 336. Washington, D.C.: WJrld Bank, 1979, p. 35; and Turnham, David, The Ellployrrent Problem in Less Developed countries. Paris: O.E.C.D., 1970, pp. 58-60. 21. The following analysis oovers the job-seeking unalployed only, i.e., tOOse who did not have a job during the reference week rot who were seeking one. 22. Bairoch, P., Urban Un Countries: The Nature of the Problem and Prop:>sals for Its SO ution. Geneva: ILO, 1974; and Turnham, D., The Errployrrent Problem in Less Developed COUi1tries, p. 76. 23. Todaro, Michael P., "Incane Expectations, Rural-Urban Migration and Enployrrent in Africa," in Ellploynent in Africa: Sorte Critical Issues. ~eva: ILO, 1977, pp. 43-69; and Standing, Guy, Lal:our Fbrce Participa- ~, pp. 207-11. 14