Does Labour Supply Respond to Globalisation? Malaysia Evidence from Micro Data

Transcription

1 Does Labour Supply Respond to Globalsaton? Malaysa Evdence from Mcro Data Poo Bee Tn, Rahmah Ismal and Norasmah Othman Abstract Globalsaton s a phenomenon that cannot be avoded. As globalsaton allowed free flow of nputs ncludng labour, t may affect job opportuntes for the locals. Therefore, nvestgate the determnants of labour supply s essental n understandng the structure of labour market n the new era of globalzaton. The objectve of ths artcle s to examne labour supply by takng nto account the globalsaton effect. The study covers 3885 households n Pennsular Malaysa who are chosen usng stratfed random samplng. The labour supply model wll be the bass for the analyss. The basc labour supply determnants are own wage and non-labour. However, the extended labour supply model ncorporates other varables lke spouse wage,number of chldren and ndvduals characterstcs lke educaton level and age. Besdes, the globalzaton ndcator wll also be ncorporated as another ndependent varable. Keywords globalzaton, head of households, labour supply, wage I I. INTRODUCTION N the recent debate on the effects of ncreasng nternatonal ntegraton on the labour market, most of the attenton has been devoted to evaluate the mpact of trade on wages and employment. However, there mght be other paths through whch globalzaton nfluences the labour market, one of these s the effect on labour supply. Globalsaton s a process where nterlnk between countres n the world becomes more ntense and flow of nputs between one country to another wll be much easer. In the context of labour market, the nflow of labour nput s more relevant, because t gves mplcaton on local labour especally n terms of job opportuntes. Indvdual percepton on the mpact of globalzaton may change ther atttude towards beng workng, but on the other hand, the labour supply may ncrease to cope wth ncreasng cost of lvng due to globalzaton. Labour supply plays a very mportant role n an economy s development. A robust and suffcent labor force promotes development, and development, n turn, feeds back on labour market condtons. Two aspects of labour supply have been mportant; frstly, quantty of labour as represented by populaton growth rates, Poo Bee Tn s wth School of Economcs, Faculty of Economcs and Management, Unversty Kebangsaan, Malaysa, Bang, Selangor, Malaysa ( phone : ; emal: pbt@ukm.my). Rahmah Ismal s wth School of Economcs, Faculty of Economcs and Management, Unversty Kebangsaan, Malaysa, Bang, Selangor, Malaysa ( phone : ; emal: rahs@ukm.my). Norasmah Othman., s wth the Faculty of Educaton, Unversty Kebangsaan Malaysa, Bang, Selangor, Malaysa (phone: ; e-mal: ln@ukm.my). rsng female labour force partcpaton rates and mgraton. Secondly, qualty of labour as represented by educaton levels and health status (lfe expectancy). In Malaysa, labour supply or labour force refers to those who, durng the reference week, are n the years age group (n completed years at last brthday) and who ether employed or unemployed. The most mportant determnant of labour supply s labour force partcpaton rate (LFPR), whch s defned as number of labour force dvded by number of populaton aged years old. In 2010 total number of labour force n Malaysa was 11, thousand persons or abour one-thrd of Malaysan total populaton. Of ths, 11,171 thousand persons are employed and the remanng thousands are unemployed. The LFPR for the total economy was declnng from 64.9 percent n 2001 to 63.2 percent n 2007 and the same patterns are shown by male s and female s LFPR [1]. Our startng pont s the tradtonal vew of labour supply determnants such as own wage, spouse wage, non labour, number of chldren, age and educaton level. Nowadays, t s mportant to take nto account effect of globalzaton because globalzaton allowed free flow of nputs ncludng labour, whch may affect job opportuntes for the locals. Consequently, nvestgate the determnants of labour supply s essental n understandng the structure of labour market n the new era of globalzaton. Ths study contrbutes to the lterature by nvestgatng the tradtonal determnants of labour supply as well as the effect of globalzaton to labour supply. In addton, studyng the behavor of labour market can gve rse to mportant polcy mplcatons. Therefore, the objectve of ths artcle s to examne determnants of heads of household labour supply by takng nto account the globalzaton effect. The analyss wll be based on the data collected from the feld survey n The study covers 3885 households n Pennsular Malaysa who are chosen usng stratfed random samplng. The data nclude nformaton on heads of households, ther spouses, famles, educaton background and employment background. Snce the man purpose of the study s to look at the mpact of globalzaton on labour market structure, the data also cover questons on globalzaton from the pont of vews of respondents. II. LITERATURE REVIEW The elastcty of labour supply wth respect to wage rate plays a crtcal role n many economc polcy analyses. There are many studes of labour supply elastcty accessble. Most of the emprcal results for the elastcty of hours of work wth 1851

2 respect to the wage rate sgnfcantly dffer n sgn and range. It appears from the lterature that the frst estmaton on the labour supply elastctes was made by [2] n hs Theory of Wage. He collected and aggregated the data for 38 US ctes from census of manufacture and examned both tme seres and cross-secton data on hours of work and hourly earnngs. He concluded that labour supply elastctes are between negatve 0.1 and 0.2. [3] mentoned that modern labour supply often separates the and substtuton effects and make use of mcro data nstead of aggregated data. Usng data from US coal mnng n the frst decades of the 20th century, [4] fnds the labor supply elastcty to be n the range n the short run and nfnte n the long run. However, [5] shows that the quanttatve relatonshp between employment and wages depends crucally on whether wages are regressed on employment or the other way around, and ndcates that the reason s measurement error. He concludes that even though t s reasonable to nterpret ths relatonshp as evdence of upward slopng supply curves, such regressons are just not very nformatve on the supply. [6] report that across estmates of own wage labor supply elastctes n varous studes; the medan elastcty was 0.08 for men and 0.78 for marred women. For cross wage elastctes, [7] pont out that a medan spouse wage elastcty of 0.13 for marred men s labor supply and for marred women s labor supply, although study of the 1980s by [8]analyzng labor supply condtonal on havng postve hours, reports a cross elastcty of roughly -0.4 to -0.5 for women and to for men. These surveys ndcate that women s labor supply s consderably more senstve to ther own wages than s men s. Ths dfference s usually explaned by the tradtonal dvson of labor n the famly, n whch women are seen as substtutng among market work, home producton and lesure, whle men are vewed as substtutng only or prmarly between market work and lesure [9]. Most of the studes about effect of spouse s wages to labour supply focus on labour supply of wves respond to husbands wages. Snce, gven tradtonal gender roles, women are perceved as secondary earners wthn the famly, ther labor supply s lkely to be more negatvely affected by ther spouse s wages. [10] examned the effect on a marred woman s labour supply decson of non-labour and of her own wage rate n Hungary. The mcro analyss showed that total montly household non-labour has been defned as the sum of two dfferent components, frst, the sum of sosal transfers that are receved by any member of the household and second, the montly share of yearly profts. Frstly, the emprcal results ndcated that wage elastcty for marred woman s estmated to be sgnfcant postve. Secondly, household earnngs other than the wve s (alternatvely, earnngs of husband when he s head of the household) were estmated to have no sgnfcant negatve effect on probablty of supplyng labour). These mply that Hungaran women take ther labour supply decsons ndependently of ther husband s or other members of the household s earnng. Fnally, the non-labour effect s consstent wth lesure beng a normal good. On the other hand, [11] demostrated that the labour supply of Canadan wves responded strongly changes n husband s wages durng the 1980s. Generally, as workers age, they may prefer to decrease ther number of workng hours due to health constrants or care oblgatons. As a result, one may expect to observe a steady drop n workng hours before full retrement. [12] concludes that among older Amercan workers who left ther job between 1992 and 2000 about 13% would have stayed n ther job f they could have reduced ther number of workng hours. In Sweden, about 7 percent of the workers aged 50 years and above clam to have physcal problems whch restrct them from contnung work n the present occupaton untl the offcal retrement age, but that shortenng workng hours would solve the problem [13]. Educaton s sgnfcantly correlated wth economc growth, whch, n turn, affects the labour supply patern. [14] found a sgnfcant postve effect of educaton on the female labour supply to the urban labour markets n Sudan. Conversely, [15] showed that ncreases educaton levels are assocated wth decreases n household s labour supply and ncreases n ther off-farm labour supply n rural Ghana. [16], who studed labour supply n Sudan, found a negatve and sgnfcant relatonshp between educaton and labour supply n the agrculture sector. Although the emprcal fndngs are rather vared, a strong research tradton supports human captal theory as a theoretcal framework to clarfy and predct the relatonshp between educaton and labour supply. Supposedly, labours wth hgher educatonal levels are more lkely to be actve n the labour force snce educaton s an nvestment that s postvely correlated wth earnngs potental. Most of the lteratures on labour supply gve specal attenton on females because of ther dfferent characterstcs compared wth males especally when dealng wth house chores. Most studes mentoned about a strong relatonshp between females labour supply and famly sze ncludng the age structure of the chldren. The relatonshp between chldren s age and female labour supply was the man focus n the studes by [17]-[18] and [19]. All of them demonstrated that chldren s age structure has sgnfcant mpact on female labour supply wth negatve effect for younger age chldren (<6 years old) and postve effect for the older age (>12 years old). The study by [20] n Mexco and [21] n the Unted States supported that mother substtute rased female LFPR. In a more specfc study, [22] found that when number of chldren ncreased by three folds, female LFPR would decrease by 8-10 percentage ponts. The frst chld s shown to gve a greater mpact on female labour supply due to hgher attenton gven to hm/her by the parent. For example, [18] found that the twn frst brth reduced female labour supply by 37 percentage ponts for age group and 10 percentage ponts for the age group. The present of chldren aged less than 6 years old has greater negatve mpact on female labour supply (see for example, [23]). [24] compared estmates of male labor supply from tme-use data wth those 1852

3 from conventonal survey data usng a so called double-hurdle model. Referrng to prevous studes they noted that the presence of young chldren normally decreases work hours for women whle the effect for males has typcally become nonsgnfcant or weakly postve. In Malaysa, [25] found that number of chldren aged below 6 years old and 7-19 years old have negatve mpact on female hours of work n the handcraft ndustry. The effects of foregn workers are tradtonally vewed n terms of complementarty or substtutablty wth natves n the producton of household servce. In the lterature revew, most of the smple theoretcal models of labour supply suggest that an ncrease of foregn workers n the natve labour market may result n lower wages and/or hgher unemployment of natves f they are perfect substtutes to mmgrants. In addton, emprcal studes typcally conclude that mmgraton has economcally rrelevant or no effects on wages and employment of natves, see [26] for survey, s that foregn workers do not have a szeable and sgnfcant effect on employment and wages of natves n the same segment of the labour market, even when the foregn workers supply shock s large [27] uses 1990 census data to study the effects of mmgrant nflows on Unted State labour market. He found that mmgrant nflows over the 1980s reduced wages and employment rates of low-sklled natves n Mam and Los Angeles by 1-3 percentage ponts. These fndng mply that massve expanson of mmgrant may have sgnfcantly reduced employment rates for younger and less-educated natves n both ctes. [28] analyss ndcates that mmgraton lowers the wage of competng workers: a 10 percent ncrease n supply reduces wages by 3 to 4 percent. Usng German data for the perod , [29] concludes that the drect mpact of mmgraton on natve wages s small as a ten percent ncrease n labor supply stemmng from mmgraton s predcted to reduce wages by less than one percent, wth a stronger negatve mpact for low-sklled natves. In recent work based on US census data, [30] extends the structural modelng approach of [28] to assess the overall mpact of mmgraton on wages whle allowng for mperfect substtutablty between natve and mmgrant workers. Ther emprcal estmates pont to a negatve, but small, drect partal effect: an mmgraton shock that ncreases the labor force n a partcular skll cell by ten percent reduces wages of natves of the same group by approxmately one percent. However, [31] argue that ncreased specalzaton mght explan why many emprcal analyses of the mpact of foregn workers on wages and employment for less-educated natve born fnd small effects. They found that foregn workers specalzed n occupatons that requred manual and physcal labour sklls whle natves specalzed n jobs more ntensve n communcaton and language tasks. Whle [32] showed that mmgraton n Italy had a dsplacement effect on low educated natves (both for male and females). III. THEORETICAL FRAMEWORK AND METHODOLOGY Households are supplers of labour. Indvduals are assumed to be ratonal and seekng to maxmze ther utlty functon. The statc labour supply theory assumes each ndvdual has a quas-concave utlty functon ([6] and [7]): U = f ( C, L) (1) Where C s the consumpton and L s the lesure hours, however, ndvduals are constraned by the workng hours avalable to them. Therefore, hours of work (H) are H = T L, T s the total tme avalable. Suppose P s the prce of goods and servce, W s hourly wages rate and non-labour, Y. The ndvdual budget constrant s: PC = WH + Y PC = W ( T L) + Y PC + WL = WT + Y (2) In statc model, non labour, Y s typcally the sum of two components: asset and other unearned. The rght sde of equaton 2 often defned as full from whch consumer purchases consumpton goods and lesure [6]. Dervaton of ndvdual s labour supply functon s derve by maxmze utlty functon subject to the budget constrant. The ndrect utlty representaton of preferences s gven by: V = ( P, W, Y ) = MaxU ( C, T H ) s.t. PC = WH= Y Set the Lagragan expresson; L = ( C, H, λ ; P, W, Y ) = U ( C, T H ) λ( PC WH Y ) The frst order condtons are; L = U C ( C, T H ) λ P = 0 C L = U L ( C, T H ) + λ W = 0 H L = PC + WH = + = 0 λ Or smply the frst two condtons take the famlar form; U L ( C, T H ) W = = Margnal Rate of Substtuton U ( C, T H ) P C Therefore, the ndvduals labour supply equaton s obtaned as below; H = H ( P, W, Y ) or W H = H (, Y ) P H = H ( HSW, Y ) (3) W Where = HSW P We assumed that P s constant, thus nomnal wage ( W ) s equal to the real wage (HSW ). Besdes these two basc varables, there are other factors that could determne 1853

4 ndvduals labour supply such as spouse wage, number of chldren and ndvduals characterstcs lke educaton, age, and so forth that can be summarzed as Z. Therefore, equatons (3) can be wrtten as, H = H ( HSW, Y, Z1, Z 2 ) (4) Table I shows results of the relablty test for plot study data. The questons cover some statements to measure globalsaton ndcators. All values of Cronbach Alpha are above 0.8, whch are consdered as very good. Ths ndçâtes that all constructs are approprate n measurng globalsaton. Therefore, no modfcaton were made on the questonnares after the plot test. IV. MODEL SPECIFICATION In order to acheve the objectve of ths paper, the study uses the logstc regresson model for labour supply equaton to capture head of households and does not go to work and who go to work. In general, logstc regresson s used to measure the functonal relatonshp between the qualtatve dependent varable and the quanttatve and qualtatve TABLE I RELIABILITY TESTS Gloabalsaton Indcators Determnants of choces for chldren s educaton before 1995 (23 constructs) after 1995 (23 constructs) Trend of demand for hgher educaton before 1995 (12 constructs) after 1995 (12 constructs) Importance of feld of chldren s educaton before 1995 (11 constructs) after 1995 (11 constructs) Importance of chldren s educatonal nsttutons before 1995 (19 constructs) after 1995 (19 constructs) Influence of globalsaton on chldren s aqured sklls before 1995 (10 constructs) after 1995 (10 constructs) Globalsaton effect on career development before 1995 (6 constructs) after 1995 (6 constructs) ndependent varables. The dependent varable wth a dchotomous character s used, whereby, the value 1 denotes f a head of household s workng and 0 values denotes f a head of household s not workng. Therefore, the model s estmated n logstc bnomal form. Model used s as follows: z P = 1(1 + e ) (5) wth P as the probablty of workers havng been moble ( Y = 1). Probablty to choose the other s ( Y = 0 ) wrtten as; z Cronbach Alpha (N=30) (1 P ) = 1 /(1 + e ) (6) Therefore, the probablty of a worker changng jobs s; Impact of globalsaton on wages and employment Source : Plot Survey 2011 z e = P /( 1 P ) (7) The model s then transformed to a logarthm model to produce the equaton; z z = ln( P /(1 P ) = ln = e = β 0 + β 1 X 1 + β 2 X β n X n (8) The estmaton of the logstc model utlses lkelhood rato test (LRT) as ndcator for ftness of the model and t-test for dentfyng the sgnfcant of the parameters. The estmaton model for ths study s as follows, LS = β 0 + β 1HSW + β 2 SW + β 3HEDU + β + 4 HAGE + β 5 NLY β HFL + 9 NUMC β 7 GLOB β β HWEG + µ (9) Where, LS s head of households labour supply denote 1 f workng and 0 otherwse, HSW s head of households monthly wages, SW s s spouse monthly wages, HEDU s head of households level of educaton, HAGE s head of households age, NLY s household non- labour, NUMC s household number of chldren, GLOB s effect of globalzaton measured by household s mn score based on percepton, HFL s dummy varable for the present of foregn labour n head of households job place, HWEG s head of household percepton of wage effect from the present of foregn labour, µ s error term and I s household. The level of educaton s categorzed nto two measures, year of schoolng and level of educatonal attanment (secondary and tertary). V. RESULTS AND DISCUSSION Table II presents head of households profle from the survey data. In total, 3885 head of households were ntervewed. In terms of demography profle, expectedly, the majorty of the head s of the household are men (67.0 percent). Malay make up the largest populaton of the sample (69.5 percent), follow by Chnese (27.1 per cent), Inda (2.2 percent) and other races (0.6 percent). The majorty of the head of the households who are n the range between year of age, whch represents 33.4 percent of the sample. Very small percentage of the households s aged above 56. A greater proporton of the head of the households are attanng secondary level of educaton (48.5 percent), than are attanng degree (19.1 percent), dploma (18.7 percent) and prmary level of educaton (12.6 percent). As shown n Table II, more than half of the respondents (55.2 percent) work n the domestc prvate sector, 32.7 percent n publc sector and only 6 percent n multnatonal prvate sector. In 2011 (year of data collecton), sector breakdown of respondents was as follows : servces 64.8 percent, manufacturng 11.3 percent, agrculture 10.9 percent, constructon 6.5 percent and mnng 0.3 percent. The 1854

5 majorty of them are servce and sales workers (22.7 percent), professonal (16.9 percent) and clercal support workers (9.4 percent). The respondents were also asked to choose among seven ranges of monthly. Of the total number of 3885 respondents, the majorty (36.9 percent) of the respondents ndcated monthly of RM1001-RM2500, 28.3 percents receve monthly of RM2501-RM4000. The head of households who receve monthly RM8001-RM10000 and more than RM are very few wth the percentage of about 2.5 percent and 2.3 percent respectvely. In terms of non-labour, majorty of the head of the households (98.2 percent) receve less than RM1000 per month. Descrptve statstcs of the explanatory varables are shown n Table III. Expectedly, the mean of the monthly for the head s of the household (RM3216) s hgher than that ther spouse s monthly (RM1897). In terms of educaton level, the mean of the level of educaton for head of the households s 11 years. A descrptve statstc result also reveals that majorty of the household attanng secondary educaton. On average, the head of the households are 44 years (M=44.48, SD=10.86) wth the eldest s 89 years and the youngest s 21 years. On average, respondents nonlabour monthly are RM76 (M= 76.34, SD= ) wth the maxmum RM70100 and mnmum zero. Durng the data collecton, on average, there are 3 chldren present n each Malaysan famly. As human captal s as mportant as physcal captal, the labour market structure (demand and supply) may also change whenever n the globalzaton, labour s rather moble. Therefore, the awareness of globalzaton characterstcs s chosen based on the head of the household s percepton on the effect of globalzaton on career development, wages and employment. Based on these varables, the effect of globalzaton s classfed nto two groups, () 1 to 3.99 as low ; () 4 to 7 as hgh. As shown n the Table III, the overall heads of household percepton on the effect of globalzaton on career development s hgh (M= 5.49, SD=1.16) whle the percepton of wage effect from the present of foregn labour s low (M= 1.53, SD=2.33). The logstc regresson model estmaton shows only two varables are sgnfcant to nfluence the workers decson to work n model 1, namely, head of household age and nonlabour as noted n Table IV. Varable of head of household age shows to have negatve mpact on the workng decson and sgnfcant at 1%, but non- labour has a Varable Sex male female Race Malay Chnese Inda others Age >56 Educaton Level prmary secondary dploma/stpm degree others Workng Experence <10 years years years >30 years Mssng Job Sector publc domestc Prvate multnatonal Prvate mssng Employment Sector servces manufacturng agrculture mnng constructon mssng Occupatonal Classfcaton managers professonal techncans and assocate professonals clercal support workers servce and sales workers sklled agrcultural,forestry and fshery workers craft a and related traders workers plant and machnery-operators and assemblers elementary occupatons mssng Monthly Income/Wages < >10001 Monthly Non-Labour Income < >10001 TABLE II RESPONDENTS PROFILES Frequency (N= 3885) Percentage (%) Source: Plot Survey

6 Varables heads of household monthly /wages TABLE III DESCRIPTIVE STATISTICS Heads of Household Mean Mn Max Std Devaton spouse s monthly /wage year of schoolng secondary educaton tertary educaton age non-labour number of chldren globalsaton mean (after 1995) mpact of globalsaton on wages foregn workers Source: Plot Survey 2011 Notes: * sgnfcant level 10 % *** sgnfcant level 1 % TABLE IV RESULTS OF LOGISTICS REGRESSION ESTIMATES FOR MODEL 1 (YEARS OF SCHOOLING ) Varables Model 1 (Years of Schoolng) β Exp(β) Margnal effect ntercept (0.026) head of household (0.000) spouse (0.000) years of schoolng (0.026) head of household age (0.011)*** household non-labour (0.001) * household number of chldren (0.040) globalsaton percepton (after 1995) (0.063) foregn labour (0.338) wage effect from foregn labour (0.066) Nagelkerke R N 3885 Source: Plot Survey 2011 Notes: * sgnfcant level 10 % *** sgnfcant level 1 % postve relatonshp and sgnfcant at 10%. Ths s also shown by the odd rato of less and greater than one respectvely. In model 2 (Table V), besdes the age and non-labour reman sgnfcant, another sgnfcant varable s level of secondary educaton. The log odd of beng n the labour market s hgher for head of household wth secondary level of educaton as compared to those wth the prmary level of educaton. TABLE V RESULTS OF LOGISTICS REGRESSION ESTIMATES FOR MODEL 2 (LEVEL OF EDUCATION ) Varables VI. CONCLUSION Model 2 (Level of Educaton) β Exp(β) Margnal effect ntercept (0.793) head of household (0.000) 0 spouse (0.000) secondary educaton (0.186) * tertary educaton (0.269) head of household age (0.011) *** household non-labour (0.001) * household number of chldren (0.040) globalsaton Percepton (after 1995) (0.063) foregn labour (0.340) wage effect from foregn labour (0.066) Nagelkerke R N 3885 Source: Plot Survey 2011 Notes: * sgnfcant level 10 % *** sgnfcant level 1 % The results from the both logstc labour supply models demonstrate an expected sgn for coeffcent of age, whch s negatve and sgnfcant. These mply that the older s the head of the household, the decson to work s less and ths fndng accord wth the theory. Another nterestng fndng s that non-labour s postve and sgnfcantly affects the labour supply, whch means that non-labour s one of the man concerns for the head of the household on the workng decson. But the unexpected results derved from the own wage, whch demonstrates that the decson to work among the head of households are less when wages ncrease. These fndngs contradct wth the theory and one explanaton for the contradcton s that backward labour supply curve. However, the year of schoolng varable, whch s assumed to be postvely related to labour supply, s not statstcally 1856

7 sgnfcant. But the hgher s the level of educaton for the head of the household, the hgher would be ther labour supply, whch demonstrates that the decson to work among the head of households wth secondary level of educaton are more than that those who attan only prmary level. Globalsaton s not sgnfcantly affectng the head of household labour supply, whch mean that globalsaton process s not a man concern of the head of household n securng ther jobs. Ths fndng support our hypothess that globalsaton wll play a mnor role snce the ntensty of globalzaton n Malaysa s stll at the moderate level. Other varables seem not sgnfcantly affectng head of household labour supply n the sample. ACKNOWLEDGMENT The authors wsh to acknowledge the fnancal support provded through the Research Unversty Grant Scheme (UKM-AP-CMNB ), by Unverst Kebangsaan Malaysa. REFERENCES [1] Labour force survey report, Department of Statstcs, [2] P. Douglas, Theory of wages. New York: Macmllan, [3] M. Evers, R.D. Moojj, and D.V. Vuuren, The wage elastcty of labour supply: A synthess of emprcal estmates, De Economst, vol.156, pp , [4] W. M. Boal and M. R. Ransom, Monopsony n the labor market. Journal of Economc Lterature, vol. 35, pp , 1997.lcaton), IEEE Trans. Antennas Propagat., to be publshed. [5] A. Mannng, Monosony n moton: Imperfect competton n labour market. Prnceton, NJ: Prnceton Unversty Press, [6] R. Blundell, and T. Macurddy, Labour supply: A revew of alternatve approaches, n O. Ashenfelter and D. Card (Eds), Handbook of Labor Economcs, vol. 3A, Ch 27, Amsterdam, North Holland, [7] M.R. Kllngsworth, Labour supply. New York: Cambrdge Unversty Press, [8] P.J. Devereux, Changes n male labour supply and wages, Industral and Labour Relatons Revew, vol. 56, no.3, pp , [9] J. Mncer, Labor force partcpaton of marred women: A study of labour supply. In Aspects of Labor Economcs, Gregg, H., Lews, H.G., Eds. Prnceton NJ: Prnceton Unversty Press, [10] C. Saget, The determnants of female labour supply n Hungary. Economcs Transton, vol. 7, no. 3, pp , [11] R. Morssette, and F. Hou, Does the labour supply of wves respond to husbands wages? Canadan evdence from mcro data and grouped data, Canadan Journal of Economcs, vol. 41, no. 4, pp , [12] R.G. Penner, P. Perun, and E. Steurle, Legal and nsttutonal mpedments to partal retrement and part-tme work by older workers. Washngton D.C.: The Urban Insttute, [13] E. Wadensjo, Work and penson n Sweden. European Papers on the New Welfare: The counter-ageng socety. Mlan: The Rsk Insttute, 2008, no. 9, pp [14] N.A. Maglad, Female labour supply n Sudan Narob: Afrcan Economc Research Consortum, 1998, Specal Paper 30.C. J. Kaufman, Rocky Mountan Research Lab., Boulder, CO, prvate communcaton, May [15] D. Jollffe, The mpact of educaton n rural Ghana: Examnng household labour allocaton and returns on and off the farm. Journal of Developng Economcs, vol. 73, pp , [16] O.M. Babkr and B.I. Babker, The determnants of labour supply and demand n rrgate agrculture: A case study of the Gezra Scheme n Sudan. Afrcan Development Revew, vol. 19, pp , [17] R. Gronau, The effect of chldren on the housewfe s value of tme. Journal of Poltcal Economy, vol. 81, no. 2, pp , [18] M. Rosenzweg, and K.I. Wolpn, Lfe cycle labour supply and fertlty: Evdence from household models. Journal of Poltcal Economy, vol. 88, pp , [19] T.P. Schults, Testng the neoclasscal model of famly labour supply and fertlty. Journal of Human Resources, vol. 25, pp , [20] R. Wong, and R. Levne, The effect of household structure on women s economc actvty and fertlty: Evdence from recent mother n urban Mexco. Economc Development and Cultural Change, vol. 41, pp , [21] M. Tenda, and J. Glass. Household structure and labour force partcpaton of Black, Hspancs and Whte mothers. Demography, vol. 22, pp [22] T.P. Schultz, The Influence of fertlty n labour supply of marred women: Smultaneous equaton estmates, Research n Labour Economcs, vol. 2, pp [23] E.L. Lehrer, The mpact of chldren on marred women s labour supply: Black-Whte dfferentals revsted. Journal of Human Resources, vol. 27, no. 3, pp , [24] P.S. Carln and L. Flood, Do chldren affect the labour supply of Swedsh men? Tme dary vs. survey data. Labour Economcs, vol. 2, pp , [25] Rahmah Ismal & Fatmah Sad, Women s educaton and household actvtes, n women and work, challenges n ndustralzng natons, Mamunah Ismal and Amnah Ahmad, Eds. London: Asan Academc Press, [26] G. Borjas, The economcs of mmgraton. Journal of Economc Lterature, vol. 32, pp , [27] D. Card, Immgrant nflows, natve outflows and the local labour market mpacts of hgher mmgraton, Journal of Labor Economcs, vol. 19, pp , [28] G. Borjas, The labour demand curve s downward slopng: Reexamnng the mpact of mmgraton on the labour market, Quarterly Journal of Economcs, vol.118, pp , [29] H. Bonn, Wage and employment effects of mmgraton to Germany: Evdence from a skll group approach, IZA Dscusson Paper No.1875, [30] G.I. Ottavano, and G. Per, Immgraton and natonal wages: clarfyng the theory and the emprcs. NBER Workng Paper 14188, Natonal Bureau of Economc Research, [31] G. Per, and C. Sparber, Task specalzaton, mmgraton, and wages. Amercan Economc Journal, 1(3), , [32] S. Mocett, and C. Porello, How does mmgraton affect natve nternal moblty? New evdence from Italy. Regonal Scence and Urban Economcs, vol. 40, no. 6, pp ,