Chapter 6: Livelihoods, Labor Markets, and Rural Poverty

Transcription

1 Chapter 6: Livelihds, Labr Markets, and Rural Pverty Pverty is nt having anything...nt having wrk, there is n land. Ladin Villager, L2 (QPES) While the last chapter discussed the general imprtance f ecnmic grwth in reducing pverty frm a macr perspective, this chapter attempts t examine mre clsely the livelihds f the pr frm a mre micr (husehld-level) perspective, with a view t infrming plicy-makers f ptential paths fr prmting pprtunities fr the pr. Indeed, incmes, livelihds and pprtunities are amng the tp cncerns ranked by Guatemalan husehlds. 1 In fact, despite prgress in ther areas, a significant share f husehlds d nt perceive imprvements in welfare since the Peace Accrds, and they attribute this primarily t a lack f pprtunities (factrs that directly affect their wallets, as discussed in Chapter 2). As such, imprving livelihds, particularly fr the rural pr, presents ne f the main challenges fr bth the pverty and peace agendas in Guatemala. In this cntext, this chapter begins with a review f incme surces in Guatemala. Since labr is the main prductive asset fr the pr, the chapter then examines the cnstraints faced by the pr in generating incmes, and hw such cnstraints may exclude them frm participating in the verall ecnmic system. In additin, given that these cnstraints are directly cnnected t a lack f access t credit, insurance and pprtunities, it is imprtant evaluate hw these failures relate t vulnerability and exclusin f peple r specific grups. Finally, since pverty in Guatemala is highly cncentrated in rural areas, the chapter analyzes issues pertaining t rural livelihds, including land, agriculture and nn-farm pprtunities. The chapter is mainly based n an analysis f data frm the ENCOVI, 2 thugh mst f the findings are als cnfirmed in the QPES. INCOMES AND INEQUALITY Incmes in Guatemala are unequal bth in their distributin and in their surces. As discussed in Chapter 2, incme inequality in Guatemala is quite high, with a Gini cefficient f 57. While the prest quintile f the ppulatin receives nly 3% f ttal incme in Guatemala, the tp quintile captures 62%. Disparities are als evident in the surces f incme (Table 6.1). Specifically: The pr are largely dependent n agricultural incme, reflecting their predminantly rural lcatin. Agriculture accunts fr abut half f the ttal incme f the prest quintile, as cmpared with just 3% fr the tp quintile. As discussed belw, the pr, particularly the extreme pr, seem t lack pprtunities utside the agricultural sectr, which generates relatively lw incmes. Mrever, agriculture (particularly day-labr jbs) ffers few labr benefits, such as jb security r pensins. In cntrast, the nn-pr have access t a much mre diversified set f emplyment pprtunities, with significant emplyment in nn-agricultural sectrs, such as services. As discussed in Chapter 5, grwth rates have been significantly higher in these sectrs than in agriculture ver the past 20 years. Transfers, bth public and private, are imprtant surces f incme fr the pr. Private transfers in the frm f remittances cnstitute mre than 20% f per capita incme fr husehlds that receive them. In additin, internatinal remittances (especially frm the United States and Mexic) are n average twice as large as dmestic (see Bx 6.1). Simulatins f the effects f current adverse shcks in the cffee industry (which is already affecting dmestic remittances via the decreases in seasnal emplyment) and the glbal ecnmic slwdwn (especially in the United States), suggest a sharp decline in per capita incme and a slight increase in bth verall and extreme pverty (as discussed in Chapter 5). On the ther hand, while public transfers 49

2 represent a higher share f incme fr the pr (Table 6.1), they are regressive in their abslute levels. These issues stress the need fr plicies that help affected ppulatins cpe with and adjust t shcks (as discussed in Chapters 11 and 12). They als suggest that imprving targeting mechanisms while implementing plicies fr the pr and ther vulnerable grups is essential (as discussed in Chapter 11). Table Incme Surces, by Cnsumptin Quintiles Cnsumptin quintiles Ttal Incme per capita (Q) 1,429 2,408 3,487 5,064 15,503 5,578 Labr incme (%) Agricultural Salaries Frmal sectr Infrmal sectr Net incme frm prductin Nn- Agricultural Salaries Frmal sectr Infrmal sectr Own business Frmal sectr Infrmal sectr Nn-labr incme (%) Return t capital a Dnatins, gifts Remittances Private Public Pensins, indemnizacines Other b Percentages may nt add up t 100 due t runding. a As interest received was negligible, the return t capital includes: incme frm rental f equipment, rental f prperty and the interest received. b Fr example, inheritance r lttery winnings. Surce: Wrld Bank calculatins using ENCOVI 2000, Institut Nacinal de Estadística - Guatemala. Labr Market Participatin LABOR MARKETS Participatin in the labr market depends significantly n gender and educatin levels. The Guatemalan labr frce cnsists f abut fur millin peple (with an additinal half a millin children that are emplyed between the ages 7-14). Participatin is high amng men (89%) and mderate fr wmen (44%). Pr men are mre likely t participate in the labr frce cmpared t nn-pr men, while fr wmen the ppsite is true. In fact, mre educated wmen (wh tend t be nn-pr) seem t selfselect int the labr market, which is nt surprising given the higher returns t educatin fr wmen, as discussed in Chapter 7. Finally, pen unemplyment is very lw but underemplyment (based n hurs wrked) affects abut a third f the wrking ppulatin, which is cnsistent with a number f hyptheses such as exclusinary practices r lack f emplyment pprtunities. Infrmality Labr markets are characterized by a high degree f infrmality. Infrmality seems t be grwing in Guatemala. Based n available data, it is estimated that an annual average f sme 6,400 frmal sectr jbs were lst during the 1990s. 3 Indeed, the frmal sectr seems t have been incapable f absrbing the 50

3 grwing labr supply, with surplus wrkers being pushed int the infrmal sectr. 4 The ENCOVI 2000 shws that mre than tw thirds f the emplyed and three quarters f the wrking pr are engaged in the infrmal sectr. Wmen are als mre likely t wrk in the infrmal sectr than men (with 71% f female wrkers engaged in the infrmal sectr, as cmpared with 62% f male wrkers). Imprtantly, the infrmal sectr is very dynamic and hetergeneus in Guatemala, ranging frm small scale farmers t textile wrkers and merchants (Figures 6.1 and 6.2). 5 The structure f the infrmal sectr als differs significantly between rural and urban areas. This diversity implies a variety f paths ut f pverty. Fr example, the negative crrelatin between nn-agricultural infrmal wrk and pverty, bth in rural and urban areas, highlights the imprtance f nn-agricultural emplyment pprtunities fr pverty reductin (as discussed in mre detail belw). Figure Emplyment Diversity in the Urban Infrmal Sectr, % f Infrmal Sectr Wrkers, ENCOVI 2000 Figure Emplyment Diversity in the Rural Infrmal Sectr, % f Infrmal Sectr Wrkers, ENCOVI Incme quintiles Incme quintiles Agriculture Manufacturing Cmmerce Cmmunity Other Agriculture Manufacturing Cmmerce Cmmunity Other Bx Migratin and Livelihds Migratin has lng been a dminant feature f Guatemalan life, thugh migratin patterns have shifted ver time. Seasnal Migratin. During clnial times, indigenus peple migrated t the suth cast t wrk in the prductin f indig and salt. 6 Accelerated land exprpriatins, frced labr mandates and the emergence f cffee in the late 1800s resulted in massive seasnal, and sme permanent, migratin f the indigenus frm the nrthwest highlands t satisfy labr shrtages in the suthern cffee regin, as discussed in Chapter 4. Tday, between 400,000 and 800,000 peple migrate temprarily each year. 7 Mre than tw thirds f seasnal migratin ccurs within Guatemala. Seasnal migratin is mre cmmn amng pr and rural residents, the indigenus, bilingual individuals and amng men, accrding t data frm the ENCOVI Indeed, pverty rates amng temprary migrants are high (75% as cmpared with 55% f permanent migrants and 56% f the verall ppulatin). While a lack f incme-earning pprtunities is the main mtive fr migratin accrding t migrant husehlds in the ENCOVI, migratin is als used as a cping mechanism in respnse t crises (such as the civil war r Hurricane Mitch). 8 The crisis in the cffee sectr and emerging prblems in sugar prductin will definitely affect an already vulnerable ppulatin that relies extensively n these activities fr their incmes. Internatinal Migratin. Internatinal migratin has grwn frm abut 50,000 peple in the 1980 s t mre than a millin peple in late 1990s. 9 Destinatins vary, but the main recipients are the United States and Mexic. Overall, 9% f all husehlds receive internatinal remittances, thugh a higher share f the nn-pr (12%) receive them than the pr (6%). 10 The QPES suggests that remuneratin (bth in-kind and in cash) and wrking cnditins are better in the fincas in Mexic than thse in Guatemala. Mrever, it is apparent that as cmmunity members migrate abrad, their remittances cntribute substantially t thse families that receive them. Indeed, there are bvius differences in the huses f thse with family members living in the US and thse withut. It seems that internatinal migratin intrduces a certain degree f within-village inequality at the cmmunity level due t remittances and gds sent back hme. Permanent Migratin. The ENCOVI shws that permanent migrants are mre likely t be nn-pr living in urban areas and having cmpleted primary educatin. They are als predminantly nn-indigenus. On the ne hand this may reflect limitatins faced by the indigenus in finding permanent pprtunities and relcating prir t the signing f the Peace Accrds, such that permanent relcatin was cntained amng the better ff, namely, the nn-indigenus. On the ther hand, the negative crrelatin between permanent migratin and pverty may als suggest permanent rural-t-urban migratin as a pssible path ut f pverty. While it is impssible t test this hypthesis with ENCOVI data, it may be the case that by migrating, these peple managed t take advantage f better pprtunities, assets and services in urban areas such that mst f them are nt pr tday. 51

4 Labr Plicies Active labr market interventins are nt effective instruments fr reducing pverty. In fact, existing labr market plicies d nt benefit the pr since the d nt reach them. The dminance f infrmal sectr jbs amng the pr puts pr wrkers ut f reach f such plicies. 11 In fact, data frm the ENCOVI shw that, while abut 40% f salaried wrkers earn less than legal minimum, this share rises t 60% f pr wrkers. 12 Mrever, nly 16% f the pr receive ther labr benefits (such as the thirteen mnth salary bnus). Indeed, villagers in the QPES cmmunity KA1 describe the schemes used by emplyers t avid paying labr benefits t the finca wrkers (including dismissing them regularly thrughut the year s as t maintain them n nn-permanent cntracting status and hence avid paying the benefits). Since active labr plicies intrduce distrtins int the labr market, these plicies can in fact be cunterprductive. As discussed belw, plicies shuld instead fcus n enabling the pr t participate in the labr market, which requires investments in educatin, technical training, and infrastructure (such as rads). Such investments will help break dwn barriers t pprtunities and imprve renumeratin. Opprtunities The pr have fewer pprtunities fr higher-paid jbs and a limited ability t diversify incme surces. The nn-pr are twice as likely t wrk in a higher paying jbs in the public sectr r whitecllar ccupatins as the pr. Mrever, the pr are three times as likely t wrk in agriculture as the nn-pr. In fact, prer husehlds appear t be rather cnstrained in emplyment pprtunities, dividing their labr between self-emplyment, blue-cllar jbs, and agriculture, which all yield significantly lwer incmes than ther sectrs. In additin, husehld emplyment prtflis reveal a pattern whereby prer husehlds are mre hmgenus in their ccupatins (e.g. mst f the individuals are in agricultural), while fr nn-pr husehlds there is mre diversity amng the ccupatins f wrking members. While sme f this disparity is explained by differences in human capital accumulatin, this unequal distributin in pprtunities and incme-generating pssibilities is als attributable t the lack f spatial integratin fr these grups. Nn-Spanish speakers and indigenus ppulatins face similar cnstraints in emplyment pprtunities. Gegraphic lcatin is crrelated with pverty and emplyment pprtunities. The spatial prximity t a bigger city may ffer a number f advantages t a husehld such as emplyment pprtunities but als access t services and infrastructure nt available in a smaller cmmunity. Cmbining data frm the 1994 census (which allws fr the cnstructin f municipal ppulatin sizes) with thse frm the ENCOVI reveals that lcatin is central t emplyment pprtunities. First, 75% f the husehlds that reside in small municipalities (less than 10,000 peple) are rural, whereas 40% f thse living in municipalities with mre than 30,000 peple are rural. Secnd, pverty rates are significantly higher amng husehlds in smaller municipalities. Third, the share f nn-farm incme is higher fr thse husehlds residing in larger municipalities implying that nn-farm emplyment pprtunities (that yield higher incmes) are mre likely t be available in these areas. In fact in the rural areas, the share f nnfarm self-emplyed incme fr husehlds living in larger municipalities is almst twice the incme f thse frm smaller municipalities. Therefre, if municipal size is a prxy fr pprtunities and infrastructure, these crrelatins imply that integrating husehlds with markets and the rest f the ecnmy is key fr allwing the pr t access pprtunities. Wages and Returns t Labr Real wages have fllwed a divergent trend between sectrs. The evlutin f real mnthly wages by industry ver the last decade reveals tw interesting patterns. 13 First, while in the early 1990s, real mnthly wage grwth was relatively equal amng the different industries, dispersin in wages has increased in the late 1990s, as has been bserved fr many ther cuntries in LAC. The increase in wage inequality culd be a respnse t refrms 14 and trade liberalizatin, which shifted relative wages in favr f higher skilled 52

5 jbs (and perhaps has driven the apparent increase in verall inequality, as discussed in Chapter 2). Secnd, agricultural wages have been increasing much slwer than in ther sectrs since the mid-1990s. Nnetheless, despite lagging levels, wages in agriculture seem t have increased in real terms during the 1990s, prbably explaining a gd part f the imprvements in pverty. Figure 6.3 Evlutin f Real Mnthly Wages, by Sectr 3000 Mining Cnstructin Quetzales Transprt Cmmerce Cmmunity Manufacturing Basic Services Agriculture Adjusted fr prices Surce: ILO (labrsta.il.rg) Year The lwest labr earnings ccur in agricultural ccupatins, rural areas, the infrmal sectr and amng marginal grups such as the pr and indigenus. Real hurly wages average Q Wages are mre than twice as high in urban as cmpared with rural areas and als fr the nn-indigenus as cmpared with the indigenus. In additin, the average real wage f Q3.3 in agriculture is almst five times smaller than the Q15.8 wage fr financial services. The average hurly wage in the infrmal sectr is less than half f that in frmal ccupatins in the private sectr. Wages decrease dramatically fr prer individuals and increase with higher levels f educatin. Similar patterns are als bserved fr earnings amng the self-emplyed. Men in lwer-skilled ccupatins receive mre than wmen while there is mre wage equality amng higher skilled jbs. Discriminatin in labr markets is ften reflected bth in hiring practices but als in the earnings differentials between different grups such as men and wmen. In Guatemala, wages fr men are up t 50% higher than fr wmen in jbs like manufacturing and cmmerce. This wage differential is smaller and even negligible in the public sectr r white-cllar ccupatins, where typically the educatinal attainments are higher. Yet, as the analysis shws belw, wage differentials cannt be fully explained by educatinal attainments alne, implying that there is a high degree f discriminatin. Ability t speak Spanish is crrelated with higher earnings. Men and wmen wh speak Spanish earn mre than 30% mre than thse wh d nt, even when ther factrs are taken int accunt using regressin analysis. This is als true fr bilingual speakers, indicating the rle f language ability. Yet, as shwn belw, neither educatinal attainments nr language ability are enugh t explain earnings gaps between indigenus and nn-indigenus, implying the existence f labr market discriminatin. Wage discriminatin is high fr indigenus grups. While human capital endwments explain sme f the variatin in earnings, wage discriminatin (based n the unexplained part f the determinants f 53

6 earnings) is prevalent fr the indigenus. 16 The average wage gap between indigenus and nn-indigenus wrkers is 50%. Cntrlling fr differences in human capital endwments, experience, sectr f emplyment and ther characteristics, a significant share f this gap can be attributed t wage discriminatin (95% fr men and 35% fr wmen) against indigenus wrkers. Similarly, clse t half f the wage gap between men and wmen can be attributed t wage discriminatin (cntrlling fr ther factrs), thugh the average wage gap between men and wmen is small (1.3%). Child Labr Child labr is cmmn in Guatemala. Abut half a millin children between the ages f 7 and 14 are emplyed in Guatemala, with a third f them wrking in plantatins (mainly cffee and sugar fincas). Mst f these children cme frm pr husehlds (75%) and reside in rural areas (80%). Bys are mre likely t wrk in agriculture, while girls generally wrk in bth agriculture and cmmerce. They receive significantly lwer wages than adults and wrk abut 30 hurs a week, seriusly inhibiting the ability t attend schl (see Chapter 7). RURAL POVERTY AND LIVELIHOODS As discussed in Chapter 2, rural areas have disprprtinately higher rates f pverty than urban areas. This sectin examine the livelihds f the rural pr, arguing that agriculture and land refrm are unlikely t prvide significant plicy levers t reduce pverty. Subsistence farming and traditinal exprts such as cffee, in particular, are unlikely t be significant surces f emplyment grwth fr the rural pr. Nntraditinal exprts culd serve as a ptential surce f grwth, but their reach has been limited in scpe, particularly fr the pr. Rather, data frm the ENCOVI suggest that nn-farm pprtunities are likely t drive rural grwth and emplyment pprtunities. As such, remving cnstraints faced by the pr t engage in such ccupatins is crucial. Agriculture, Land and Rural Pverty The vast majrity f the rural pr are subsistence farmers r agricultural day labrers. Sme 87% f the rural pr depend n agriculture, either as small-scale subsistence farmers r agricultural labrers (Table 6.2). Indeed pverty rates amng these grups are significantly higher than thse whse main surce f incme cmes frm nn-agricultural surces. As discussed in Chapter 3, dependence n agriculture is significantly related t lwer cnsumptin levels (even after cntrlling fr ther differences). Access t land is inversely crrelated with pverty. Land is nt nly necessary fr agricultural prductin, but can als serve as cllateral, allwing husehlds t btain credit fr input purchases r diversify their incme prtfli by engaging in ther activities. Nnetheless, land wnership in Guatemala is highly skewed and unequally distributed, as discussed in Chapter 4. The pr tend t have smaller plts and pverty rates are higher amng thse with smaller plts (Table 6.2). Indeed, larger hldings are significantly crrelated with higher cnsumptin in rural areas (even cntrlling fr ther factrs, as discussed in Chapter Table 6.2 Rural Pverty by Land Status and Main Surce f Incme % f rural ppulatin % f rural pr Pverty rate (P0) All rural Landwners hectares hectares hectares hectares >15 hectares Tenants Landless Husehlds a Agricultural day labrers Nn-agricultural wrkers a. Landless husehlds are divided int thse that derive mst f their incme frm agriculture and thse that derive mst f their incme frm nn-agricultural surces Surce: Wrld Bank calculatins using ENCOVI 2000, Institut Nacinal de Estadística - Guatemala. 54

7 3). Mrever, the relative gegraphic islatin f the pr suggests that their land is prly lcated, further depressing the values f this asset. The plts wned by the pr are als likely t be f lwer quality than thse wned by the nn-pr. In additin, pr landwners are less likely t have land titles r use land markets. Land titles allw farmers access credit by using their land as cllateral fr credit (see Bx 6.2). They als ffer security and even prvide incentives fr the farmer t be mre prductive. Overall, nly 41% f landwner husehlds pssess frmal title t their land. Only a third f the pr have land titles cmpared t almst half f the nn-pr. Larger landwners are als mre likely t have titles. Mrever, the nn-pr use land markets mre than the pr. In particular, mre than half f the land wned by nn-pr husehlds was purchased n the market as cmpared with nly a third f the land wned by extreme pr husehlds (the remainder being acquired by inheritance). Bx Rural Credit Credit can prvide an imprtant input t bth agricultural and nn-farm prfitability. Credit allws husehlds t acquire land and ther inputs. It als enables husehlds t diversify their incme surces by engaging in ther activities besides farming. Nnetheless, access t credit in rural Guatemala is minimal. Only 13% f the rural husehlds applied and received any kind f a lan (15% amng the nn-pr and 13% amng the pr). Just 10% f landless husehlds applied and received a lan as ppsed t 15% fr landwner husehlds, indicating the imprtance f cllateral (as discussed belw). Lan sizes were significantly higher fr the nn-pr than the pr. Lack f cllateral is ne f the mst frequent reasns husehlds refer t fr nt applying fr a lan. The ENCOVI suggests that abut 40% f rural husehlds wanted t receive credit but chse nt t apply. Amng thse wh believed they needed credit, the tp three reasns fr nt applying fr a lan were: (a) lack f cllateral; (b) t expensive; and (c) fear f ineligibility fr a lan. In the case f cllateral, it is nt necessarily the lack f assets per se but the lack f frmal wnership f the assets that may prevent husehlds t access credit. Fr example, lack f land titles (discussed abve) may disable farmers frm btaining credit even thugh they d wn the land. Interestingly, it appears that credit institutins d exist in rural areas. Hwever, the fact that that peple feel that lans are expensive may be a signal f a nn-cmpetitive credit market (fr example, a few lcal infrmal mney lenders). In additin, a lack f infrmatin may lead t peple with the wrng impressins abut their ability t qualify fr a lan as well as the csts and risks assciated with lending. As such, addressing these issues f infrmatin and supply is imprtant. In fact, a recent study n rural financial services in Guatemala recgnizes the shrt supply f frmal lending institutins as an imprtant impediment t btaining credit in rural areas. 17 Indeed, nly half f the husehlds that reprted receiving lans in the ENCOVI survey btained them frm frmal lenders. Furthermre, the pr and landless are mre likely t receive credit frm infrmal lenders, crrbrating the findings that frmal lenders require land titles as cllateral and the fact that the pr face cnstraints in accessing credit frm frmal institutins. Nnetheless, interventins in credit markets shuld be cmplemented by investments in rural infrastructure, educatin and infrmatin access in rder t raise living standards in rural areas Market-based land refrm effrts are prmising and shuld cntinue t be pursued. Hwever, existing pilt land prgrams have prved difficult and slw due t their high csts and ther design issues (Bx 6.3). Further, lw agricultural returns in traditinal crps require that beneficiaries f these prgrams lk fr better prspects in nn-traditinal alternatives. Hwever, the staggering csts f prviding the bulk f the rural pr with gd quality parcels and cmplementary inputs suggest that this is nt a feasible ptin t eradicate mass pverty in the rural sectr. Similarly, agriculture verall is unlikely t prvide a significant rute ut f pverty fr the bulk f the rural pr. Agriculture has faced declining grwth rates ver the past several decades, cntracting as a share f GDP (see Chapter 5). Within agriculture, traditinal crps such as cffee, which tend t emply a significant number f wrkers, are cntracting in the face f a structural terms-f-trade crisis, as discussed belw. Nn-traditinal crps have increased significantly, but nt enugh t replace the earnings and emplyment pprtunities lst by the cffee crisis, as discussed belw. As such, althugh the pr will 55

8 likely t cntinue t depend n agriculture as an imprtant surce f incme, it is unlikely that agriculture will prvide the slutin t the pverty prblem r that many peple will escape pverty via agriculture. Bx 6.3 Land Redistributin Prgrams in Guatemala Land redistributin is a sensitive tpic, bth plitically and culturally. Since the 1980s, a number f land prgrams have been at wrk helping farmers access land in Guatemala. Prgrams like the Fundación Guatemalteca para el Desarrll-Fundación del Centav (FUNDACEN), the Fnd para la Reinserción Labral y Prductiva de la Pblación Repatriada (FORELAP), and the Fnd Nacinal de Tierras (FONATIERRA) have benefited mre than 7,000 husehlds. As part f the Peace Accrds in 1996, the Guatemalan Gvernment established the Cmisión Institucinal para el Desarrll y Frtalecimient de la Prpiedad sbre la Tierra (PROTIERRA) whse respnsibilities include: (i) a cadastral-based land registry; (ii) a land fund t prmte market-driven land refrm; (iii) land cnflict reslutin mechanisms and free legal services with special attentin t land access and land traditinal management by rural cmmunities; (iv) a natinal gegraphic infrmatin system; (v) a cmprehensive land tax system; (vi) agricultural develpment; and (vii) rural investment prgrams. The idea f the land fund is t allw pr husehlds t access land via a credit subsidy. The prgram gives a lan t qualified husehlds that allws them t buy land and inputs. In additin, the participants receive technical assistance. The lan lasts fr 4 years and it csts 5% in annual interest. Up t tday, abut 5,000 husehlds have participated in the prgram. Hwever, this and previus prgrams have been criticized fr being slw and the verall target f 335,000 (in the case f the land fund) t be unfeasible. In additin, critics mentin the lack f incentives t repay the lans as repssessing the land is hard t implement, preferential treatment in the way that land is allcated, the lw interest rates charged and the prgram s pssible vercrwding f ther land prjects as imprtant prblems. Currently, bth the gvernment and lcal agencies are cnsidering alternative market-based mechanisms t facilitate land access. Fr example, rental markets are likely t be cheaper t implement. In additin, title prgrams (such as the Registr General de la Prpiedad (RGP) r pilt prgrams such as the Catastr prject may nt nly allw husehlds t access credit but als enable them t participate in rental prgrams. Finally, land leasing prgrams with the ptin t buy the land are als cnsidered. Surces: The land fund prgram, (1998); Carrera (1999) Traditinal Agricultural Exprts: Vulnerability and the Cffee Crisis The wrldwide structural change in the cffee industry is seriusly affecting Guatemala. Cffee has always played an imprtant rle fr the Guatemalan ecnmy. It is the mst imprtant exprt f the cuntry with receipts f mre then $570 millin in 2000 (20% f ttal exprt earnings), making Guatemala the fifth largest cffee exprter in the wrld. 18 Hwever, the recent entry f a number f new prducers (particularly Vietnam), as well as verprductin in Latin American cuntries (e.g. Brazil) have severely depressed internatinal cffee prices, resulting in significantly lwer revenues fr cffee prducers in Guatemala. Indeed, the Natinal Cffee Assciatin (ANACAFE) estimates that exprt vlumes in 2001 fell by 1 millin bags, t 5.3 millin, and receipts by 50% t less than $300 millin. As the dwnward trend in internatinal prices was caused by structural changes n the wrldwide cffee market, these changes are likely t be permanent (rather than a temprary price shck). Cffee prductin prvides a significant surce f incme fr many rural husehlds. The ENCOVI reveals that 11% f rural husehlds prduce cffee. Abut 140,000 rural husehlds receive incmes frm cffee prductin, f which mre than 75% are pr. 19 On average, cffee prducing husehlds received Q4,526 in cffee sales. Nn-pr cffee-prducing husehlds received almst five times mre in cffee sales than pr husehlds. Cffee incme cmprises abut 25% the ttal incme per capita, irrespective f pverty status (fr cffee prducing husehlds nly). As expected, mst cffee prducers are landwners. 56

9 Many mre peple depend n cffee prductin frm the demand f agricultural labr. Accrding ANACAFE, there are an estimated 200,000 peple permanently emplyed in the cffee industry. This figure increases t mre than 500,000 during cffee harvest seasn. Mst labrers (jrnalers) in the cffee sectr are seasnal migrants frm pr husehlds that depend n the cffee sectr t augment their incmes, as discussed in Bx 6.1. Indeed, pverty rates amng temprary migrants are high (75% as cmpared with 55% f permanent migrants and 56% f the verall ppulatin). The crisis in the cffee industry will thus affect a significant share f the rural ppulatin. Lwer revenues are likely t push sme cffee prducers t dramatically decrease their demand fr labr r frce them cmpletely ut f business. Accrding t the Ministry f Agriculture mre than 40,000 cffee prductin related jbs are expected t be lst in 2002 (ANACAFE puts this figure at 60,000). As mst f these jbs are expected t be lw-end jbs, the effect n the pr is likely t be significant. Faced with this crisis, a number plicies shuld be cnsidered. First, fr thse prducers wh will be able t survive the crisis as cffee farmers, assistance shuld be prvided t expand Guatemala s presence in well-paid niche markets (e.g., specialty, rganic, fair trade cffees) by imprving quality, helping extend lng-term cntracts and establishing cntacts with purchasing firms in develped cuntries. Secnd, fr thse wh are frced t exit the sectr, assistance shuld be prvided t help with diversificatin in agriculture and nn-agriculture, including: develping a large prtfli f substitute crps, livestck activities, service sectr activities (e.g., cmmerce, restaurants, turism, cultural activities), and light manufacturing pprtunities (e.g., handicrafts, textiles); prviding technical assistance t bst the cmpetitiveness f prmising activities; and imprving the skills f the peple displaced frm cffee. Prgrams t address the emplyment lsses als seem necessary in the shrt run, given the large number f wrkers that will likely be affected. Fr example, wrkfare prgrams (particularly seasnally-targeted schemes) culd be strengthened and expanded t prvide alternative emplyment fr thse dependent n salaries frm (seasnal) labr in the cffee sectr. Crp Diversificatin and Nn-Traditinal Agricultural Prducts Guatemala s nn-traditinal agricultural exprts sectr has grwn impressively ver the last few years. Over the past decade, exprts f varius nn-traditinal crps have increased dramatically. Fr example, exprts f fruits such as manges, papaya, berries and melns, increased frm US$14 millin in 1990 t mre than US$300 millin in Traditinal exprts, such as cffee and sugar, still dminate, hwever, increasing frm US$629 millin t US$1,044 millin ver that same perid. Nnetheless, the weakened utlk fr revenues frm traditinal exprts cmbined with the expansin f the prductin and exprt f nn-traditinal crps highlights the ptential rle f this sectr fr future pprtunities in rural areas. Hwever, the ENCOVI 2000 reveals that very few farm-husehlds prduce nn-traditinal agricultural prducts, and mst are nn-pr. The survey allws fr the divisin f husehlds amng thse that prduce exprt-related agricultural prducts (further divided in traditinal and nn-traditinal) and all ther prducts. 21 Only 23,000 husehlds prduce nn-traditinal agricultural crps as ppsed t 650,000 that prduce subsistence crps (Table 6.3). 22 Husehlds that prduce nn-traditinal agricultural prducts have better sciecnmic indicatrs, suggesting that prductin f these crps may be limited t thse husehlds that have better access t resurces. Further research culd explre the types f cnstraints that may impede farmers frm engaging in the prductin f these crps. 57

10 Table 6.3: Types f Crps Prduced, by Type f Husehld Types f crps prduced: a Bth subsistence and traditinal exprts Bth subsistence and Nn-Traditinal exprts Subsistence nly Traditinal exprts nly Husehld ppulatin (in 000 s) Indigenus (% within crp categry) Pverty (within crp categry) Extreme Pr (%) All Pr (%) Husehld distributin by land wnership (%) Tenants Landwners 0-1 hectares Landwners 1-2 hectares Landwners 2-5 hectares Landwners 5-15 hectares Landwners >15 hectares Ttal a Traditinal exprts crps are cffee, sugar, bananas and cardamm. Nn-traditinal exprts are snw peas, spruts, brccli, cauliflwer, flwers, mangs, melns, pineapple, papaya, kra and berries. Subsistence crps are the remaining crps (e.g. crn). Classificatin based n thse f the Asciación Gremial de Exprtadres de Prducts N Tradicinales (AGEXPRONT). Surce: Wrld Bank calculatins using ENCOVI 2000, Institut Nacinal de Estadística - Guatemala. Agricultural Technical Assistance Agricultural extensin and technical assistance culd serve as ptentially levers t prmte rural grwth and reduce pverty, particularly given the imprtance f crp diversificatin and the need t imprve farm prductivity in Guatemala. Nnetheless, data frm the ENCOVI shw that nly 3% f the farmers in rural Guatemala received such assistance in While mst technical assistance prvided by nn-public institutins seems t be well targeted t pr husehlds, mre than 70% f public technical assistance is received by nn-pr husehlds, suggesting inadequate targeting f public interventins. Nn-Agricultural Opprtunities: a Ptential Rute Out f Pverty? Nn-agricultural pprtunities may prvide a pssible channel fr escaping pverty. Much f the recent empirical literature n the relatinship between nn-agricultural incmes and rural pverty clearly indicates a strng relatinship between the tw. 23 First, by diversifying their incme prtflis, rural husehlds can augment their incmes and minimize adverse incme shcks frm farm activities. Secnd, as nn-agricultural incmes increase the husehlds cash liquidity, husehlds can access farm inputs easier thus raising farm prductivity. Finally, the nn-farm sectr ffers pr landless husehlds (therwise unable t engage in farm activities) an ptin fr incme generatin. Indeed, rural husehlds that d nt depend n agriculture are less likely t be pr. The ENCOVI reveals a negative crrelatin between nn-agricultural incmes and pverty in Guatemala. Landless husehlds have significantly better sci-ecnmic indicatrs than thse that depend n agriculture: lwer pverty rates (see Table 6.2), higher incmes and cnsumptin levels, lwer husehld sizes and higher levels f educatin fr the husehld head. In fact, almst half f the rural nn-pr husehlds are landless, typically wrking as self-emplyed entrepreneurs in cmmerce r manufacturing, r in nnagricultural salaried jbs (e.g., cnstructin, teaching). Only 26% f the landless husehlds are indigenus, which may be an indicatin that access t specific nn-agricultural ccupatins may be cnstrained fr indigenus husehlds. The earlier findings n indigenus labr-market discriminatin via wages als supprts the pssibility f exclusin. 58

11 Given the ptential f nn-agricultural emplyment in reducing rural pverty, it is imprtant t understand the cnstraints that specific grups face in accessing these jbs. Numerus cnstraints are plausible. Fr example, as the pr generally live in smaller and islated cmmunities (based n reginal ppulatin densities), pprtunities fr nn-agricultural jbs may be scarce. In additin, this spatial islatin implies that these cmmunities will be mre likely t lack cmplementary infrastructure (e.g. rads, electricity, telephnes). In fact, a study n the rle f basic services and nn-farm enterprise prfitability using the ENCOVI finds that: (a) the prbability f having a micr-enterprise in rural areas is significantly higher amng husehlds with cverage f mdern utilities; and (b) micr-enterprises withut access t services such as electricity and telephne cnnectins in the rural areas have significantly lwer prfits that thse wh have access. 24 Regressin analysis cnfirms that human capital as well as lack f infrastructure are bth imprtant cnstraints fr participatin in the nn-agricultural sectr. In particular, emplyment-type chice mdels indicate that lw human capital (e.g. educatin) is an imprtant impediment fr participatin in higher return ccupatins. In additin, areas where nn-agricultural pprtunities are mre wide-spread increase the prbability f being emplyed in such jbs, suggesting that the rle f lcal infrastructure and access t services is a necessary cnditin fr emplyment grwth in nn-agricultural jbs. These findings imply that while the nn-agricultural sectr culd ffer an exit path frm pverty, vercming the cnstraints that are assciated with accessing these jbs is an essential part f a rural pverty reductin plicy kit. Still mre wrk culd fcus n explicitly explring these cnstraints t fully understand the dynamics f the nn-agricultural sectr with the rest f the ecnmy. SUMMARY OF KEY ISSUES AND PRIORITIES This chapter highlights a number f key issues regarding livelihds and pprtunities fr the pr: The pr, wmen and the indigenus are cnstrained in bth emplyment pprtunities and earnings, with a high dependence n agriculture and the infrmal sectr and few labr benefits. The indigenus face significant wage discriminatin; wmen als face wage discriminatin, thugh the wage gap between men and wmen is smaller than the ne between the indigenus and nnindigenus. Gegraphic lcatin is an imprtant factr in determining bth pverty and earnings pprtunities, with smaller municipalities ffering fewer ptins. While the pr are highly dependent n agriculture (subsistence farming and agricultural jbs), agriculture is nt likely t be a dynamic surce f new emplyment pprtunities and will cntinue t shrink as a share f GDP. As such, agriculture is unlikely t serve as a majr vehicle fr pverty reductin. Similarly, althugh land is an imprtant asset, its wnership is highly inequitable in Guatemala. Mrever, the hldings f the pr tend t be: (a) quite small, ften prviding belw-subsistence incmes; (b) untitled; (c) prly lcated (gegraphically islated); and f pr quality. In additin, full-fledged land refrm is unlikely t serve as a majr vehicle fr pverty reductin due t the high csts and slw pace f land prgrams, and lw agricultural returns. Cffee, Guatemala s traditinal exprt crp, has suffered a lng-term structural price shck, further emphasizing the need fr diversificatin. This shck is likely t have substantial impacts n the pr, bth as prducers and as wrkers (permanent and seasnal). 59

12 While nn-traditinal exprt crps culd ffer ptential pprtunities fr grwth and emplyment, their scpe has been limited t date, with little invlvement f the pr. Rather, nn-agricultural emplyment culd prvide an imprtant rute ut f pverty, thugh a variety f barriers cnstrain access t these pprtunities, particularly educatin and gegraphic lcatin. These findings suggest a number f pririties fr reducing pverty and prmting pprtunities: Enhancing pprtunities shuld be at the center f the pverty agenda. A recurring pattern that arises in the analysis is the fact that the pr and ther marginal ppulatins such as the indigenus are nt able t fully participate in, r benefit frm, the verall ecnmic system. Therefre, addressing the imprvement f emplyment pprtunities is necessary. Sme specific areas fr plicy interventin that emerge frm the analysis are: Reducing the human capital gaps between the pr and nn-pr. Educatin is essential t expand the pprtunities f the pr and allw them t access higher-paying jbs. Educatin and technical training are particularly imprtant t help them access expected grwth pprtunities in nn-agricultural sectrs. Investments need t cncentrate n expanding the pr s access t educatin, as well as imprving quality s as t bst the returns t educatin (see Chapter 7). Lwering transactins csts in accessing markets. By decreasing the strng spatial disadvantage that many f pr face (especially in the rural areas), marginal ppulatins culd see dramatic increase in pprtunities via the easier access t prduct and factr markets, bth in agricultural and nn-agricultural sectrs. Thus emphasis n investments in rad infrastructure and basic services cverage is essential (see Chapters 9 and 10). Creating mechanisms t discurage labr-market discriminatin fr the indigenus and wmen. A rural develpment strategy is als key fr Guatemala s verall pverty reductin strategy. As pverty is highly cncentrated in rural areas special attentin in rural emplyment and incme generatin is imprtant. In particular, the tw main areas fr cnsideratin include: Prmting grwth f nn-agricultural sectrs, which are likely t be the main engines f rural grwth and emplyment. Despite the ptential f the nn-farm sectr as a vehicle fr reducing pverty, numerus barriers prevent the pr frm accessing such pprtunities, including disparities in educatin levels, transprt and basic infrastructure, lack f access t rural credit, and gegraphic disadvantages. Interventins shuld thus fcus n remving such barriers, with targeted investments in educatin and technical training, plicies t prmte micr-, small- and medium-scale enterprises (MSMEs) which tend t generate emplyment, and investments in basic services and transprt. Increasing agricultural prductivity and diversificatin. While agriculture is unlikely t generate enugh additinal emplyment pprtunities t reduce pverty n a large scale in the medium term, it will cntinue t be an imprtant surce f incmes fr the pr (at least in the shrt run). As such, effrts shuld be made t increase land and labr prductivity and t diversify t nn-traditinal crps. Cffee prductin shuld als take greater advantage f markets fr specialty cffees. This requires investments in human and physical capital, as well as access t new technlgies, financial institutins, and technical assistance. 60

13 Finally, safety nets and risk management are imprtant cmpnents f a pverty strategy. Even if a successful pverty reductin prgram is established, certain grups culd remain vulnerable (as discussed in Chapter 11), including: Seasnal migrants wh are already affected by the adverse structural terms-f-trade shck in the cffee industry. While they face the pssibility fr jb lss, the welfare f their families is als affected, as they depend n the remittances that they send. A strengthening and expansin f wrkfare prgrams may be needed t help supplement their incmes; Child-labrers, wh face cnsiderable trade-ffs in terms f under-investments in educatin. Explring mechanisms and incentives t mtivate tem t stay in schl is crucial since these children represent an imprtant future labr frce base; The indigenus, wh seem t have limited access t nn-agricultural pprtunities and face wage discriminatin; and Gegraphically islated husehlds, that may nt be able t take advantage f ecnmic grwth r emplyment creatin prgrams due t the high transactin csts and barriers f participating in such prgrams. 1 Surce: Wrld Bank calculatins using the ENCOVI 2000, Institut Nacinal de Estadística Guatemala. 2 See Technical Paper 1, Vakis (2002). 3 vn Hegen (2000). 4 Vn Hegen (2000). 5 Figures 6.1 and 6.2 use nly thse emplyed in the infrmal sectr lder than age 15. Other includes mining, basic services, cnstructin, transprt and financial jbs. Surce: Wrld Bank calculatins using ENCOVI 2000, Institut Nacinal de Estadística - Guatemala. 6 Ministeri de Salud (1998). 7 Estimating the magnitude f migratin is a cmplex matter. The ENCOVI 2000 estimates the number f seasnal migrant wrkers t arund 400,000 (excluding children under 7). Other surces such as Ministeri de Salud (1998) put this figure clser t 800,000. Such differences may arise frm the sample design but als frm the survey definitin f migratin. 8 The QPES and ther qualitative wrk reveal that migratin is an imprtant cping mechanism in times f crisis as well as a regular surce f earnings fr the pr. Other mtivatins fr migratin include the search fr better scial services such as educatin and health). 9 ASIES (2000). 10 Wrld Bank calculatins using the ENCOVI 2000, Institut Nacinal de Estadística Guatemala. 11 Hwever, 26% f frmal-sectr wrkers als receive less than the minimum wage. 12 The ENCOVI survey des nt distinguish between thse wh are registered in the Gvernment s minimum wage system versus thse wh are nt registered. 13 Wages have been adjusted fr inflatin using the CPI. 14 Indeed, the tw sharp changes in mnthly wages ccurred in 1997 in the basic services and transprt sectrs, and may be linked t the privatizatin prcess f electricity and telecmmunicatins industries that ccurred that year (see Chapter 9). 15 Taking int accunt all labr earnings, whether cash r in-kind, and including grss wages/salaries, the 13 th mnth salary bnus, the value f tips, etc. Wage data have been adjusted fr spatial differences using price indices. See Technical Paper 1, Vakis (2002) fr details. 16 Based n a Oaxaca-Blinder decmpsitin. 17 The Wrld Bank (1999). 18 Surce: Wrld Develpment Indicatrs 2001, The Wrld Bank. 19 It is imprtant t nte that, because the ENCOVI is a husehld survey, plantatins wned by entities ther than husehlds (e.g., crpratins, banks) are nt captured in the data cllected in the survey. The estimates presented here are husehld-based; cnclusins shuld nt be drawn abut the pverty status f all cffee prducers (e.g., crpratins, banks, etc.). 20 AGEXPRONT (2000). 21 Still, due t the survey design, these findings are nly suggestive and are nt representative. 22 Traditinal exprts crps are cffee, sugar, bananas and cardamm. Nn-traditinal exprts are snw peas, spruts, brccli, cauliflwer, flwers, mangs, melns, pineapple, papaya, kra and berries. Nn-tradable crps are the remaining crps (e.g. crn). Classificatin based n thse f the Asciación Gremial de Exprtadres de Prducts N Tradicinales (AGEXPRONT). 23 Fr a recent survey see Lanjuw and Lanjuw, See Technical Paper 7, Fster and Arauj (2002). 61