Migrant Workers Remittances and Macroeconomic Policy in Jordan. By El-Sakka, M. I. T.

Transcription

1 Migran Workers Remiances and Macroeconomic Policy in Jordan By El-Sakka, M. I. T. Dep. of Economics: Kuwai Universiy Inroducion Migran remiances play an imporan role in many labor exporing counries in he world oday. In some cases, remiances are he single mos imporan source of foreign exchange. 1 Because foreign exchange is a scan resource in many of he labor exporing counries, mobilizing remiances could ease foreign exchange bolenecks, helping hus, o miigae developmen finance, improve balance of paymens and alleviae pressures on exernal borrowing. The inflow of remiances o he labor exporing counries is, however, affeced by various macro and micro facors. The majoriy of he lieraure abou remiances focuses on he micro deerminans of he inflow of remiances. 2 Though i has been frequenly argued ha migran remiances may significanly be affeced by economic policy failures, here is relaively lile evidence abou he macroeconomic deerminans of remiances. This sudy aims a evaluaing he impac of macroeconomic policy variables on he inflow of migran workers remiances o Jordan. In conras o he previous lieraure, his sudy aims a direcly esing he effec of macroeconomic policy on he inflow of remiances. This paper is organized as follow; secion wo presens a brief background abou he Jordanian economy. The role of emigran workers remiances in Jordan in presened 1 See e.g., El-Sakka (2004), Glysos (2002), Gammelof (2002), Taylor (1999), Nisha and Bilgrami (1991), Burney (1987),. Russell (1986) 2 See e.g. Sark, 1991 and Sark and Lucas, Sark and Bloom, 1985). 1

2 in secion hree. The model of he sudy is oulined in secion four. Resuls are presened in secion five, and in secion six some conclusions and policy recommendaions are drawn. A Brief Background abou he Jordanian Economy Jordan is a relaively small counry wih a very limied resource base. I is considered as one of he poores counries by region sandards. Though i is surrounded by world s larges reserves of crude oil, i has almos none. Impors of crude oil are a major drain on he economy and on is limied foreign exchange. The counry has subsanial deposis of phosphaes and poash. Bu hese producs, as is he case for oher raw maerials, are subjec o price flucuaions and unsable demand. Jordan also suffers from a chronic waer shorage and is vulnerable o droughs, and much of is land is oo arid for agriculure. The Jordanian economy is overwhelmingly a service economy and is highly dependen on oher Arab economies, especially on Iraq. This makes he counry s foreign exchange pool very suscepible o exernal shocks, especially he ourism secor, remiance inflows, and official developmen assisance and aid from he neighboring oil rich counries. Before he Iraqi invasion of Kuwai, he Jordanian economy faced many problems including high unemploymen raes, escalaing exernal deb and declining remiances. Following he economic collapse in in he wake of a severe deb crisis, Jordan embarked upon an auseriy and resrucuring program suppored by he IMF. Pressures on he Jordanian Dinar mouned o he exen ha he counry had o devalue i currency by 50% in Oher measures were aken in he conex of an IMF s program of economic reform including governmen subsidies removal. Massive rios in April 1989 forced he governmen o keep subsidies for mos basic producs. By 1990, i was eviden ha economic performance had began o improve and Jordan was recovering from he 2

3 1989 crisis. Unforunaely his came o a hal as a resul of he decision by Iraq o invade Kuwai. The second Gulf War divided he Arab world ino wo camps. As a supporer of Iraq, Jordan was among he few counries which were deeply affeced by he invasion and is afermah. The inflow of remiances from Jordanian expariaes as well as aid from is neighboring oil rich Gulf counries were disruped as a resul. Moreover, Jordan s posiion as he main rade parner of Iraq was complicaed by he UN imposed rade sancions agains Iraq. Key Jordanian economic secors including ransporaion, agriculure and indusry, mainly serving Iraq, were severely affeced by sancions. Tourism was disruped as he number of ouriss declined by almos 25%. Jordanian expors o Arab and oher counries also declined. Kuwai and Saudi Arabia prohibied impors from Jordan because of is poliical posiion during he war. Jordanian agriculure, indusry and rade secors were hus shu-ou of he mos imporan Gulf counries markes. The crisis affeced almos all economic secors and led o a mouning unemploymen rae of nearly 30 percen and an increase in povery rae o 33 percen. During he Gulf Crisis in , some 300,000 Jordanians and Palesinians involunarily reurned back o Jordan. This compounded he effecs of he Gulf crisis on Jordan and exacerbaed he counry's already serious economic problems. However, i should be menioned ha some argue ha he long run effecs of hose reurnees are posiive because his had eased he availabiliy of skilled labor; he counry has been already suffering from is shorage, and he large ransfer of funds remied by hose reurnees (Van Hear (1995)). Under he IMF proposed program, he economy iniially recovered well wih an average growh of over 9% per year beween 1992 and Inflaion, which had exceeded 25% in 1989 fell o only 2.4%. In 1995, Jordan singed a peace reay wih Israel which inended, in principle, o esablish an Israeli-Jordanian Free Trade Area, by which Jordan would enjoy ariff concessions from Israel. I was expeced ha his move would help he Jordanian economy o carry-ou he necessary srucural adjusmen reforms in a 3

4 gradual manner in order o minimize heir economic and social coss. In realiy, boh sides agreed o lower ariffs on a limied lis of goods only. Oher indirec visible effecs were expor promoion o he Israeli marke and possibly hrough i o inernaional markes (Awarani & Kleiman (1997)). The main advanage of he peace reay was ha i had encouraged ourism, which is now growing and would in he fuure play an imporan role in he Jordanian economy. In 1996, igh moneary policy discouraged invesmen and expor-oriened manufacuring suffered from he halving of he Jordan- Iraq rade proocol, causing growh o slow down o 2.1% and inflaion rise once again o 6.5%. Jordan s small domesic marke and he dominan role of he public secor, coupled wih concerns over regional sabiliy and high lending raes have made i difficul o arac invesmen. High unemploymen and povery raes are he main problems facing Jordan. However, rade defici is also high, represening abou 25% of GDP. Finally, cenral governmen s heavy deb burden remains on of he main challenges facing Jordan in he medium erm. Since 1999, under he new leadership of Jordan, economic reform has been placed high on he agenda. However, he USA s war agains Iraq disruped economic aciviy in Jordan, bu hese effecs were shor lived. By 2004, i was clear ha economic aciviy had srenghened subsanially helped by oil grans from he neighboring Arab counries and a sandby agreemen signed wih he IMF. Economic reforms suppored by IMF aimed a (i) sabilizing he economy so as o foser growh; (ii) liberalizing rade prices; (iii) reducing public deb; and (iv) privaizing sae-owned enerprises. The governmen needs o implemen a sizable package of fiscal measures o achieve he desired fiscal defici arges (IMF 2004). The Role of Remiances in he Jordanian Economy In spie of is small indigenous populaion, Jordan plays a major role as one of he key labor exporing counries in he Middle Eas. Jordan used o expor skilled labor on a massive scale in he early sevenies of he las cenury. This had caused a serious domesic shorage of cerain skills. Serageldin e al (1981) menion ha he drain of 4

5 Jordanian emigrans reached criical levels, such ha skill shorages adversely affeced he implemenaion of Jordan s developmen plans. The Jordanian ex-crown Prince Hassan called for he creaion of an inernaional fund o compensae Jordan and oher labor-exporing naions for he negaive effecs of emigraion (Bohning (1978). Domesic wages for unskilled labor were bid up as Jordanian employers compeed for manual workers. In an aemp o replace emigrans o he Gulf, Jordan impored foreign labor from Egyp, Syria and Souh Asia. I has been claimed ha remiance ouflows of gues workers in Jordan neuralized much of he benefis of labor exporaion. Table (1) shows he officially recorded remiances in Jordan. According o he able, Jordan received some $ 30.6 billion from he inflow of emigran remiances during he period This makes Remiances he prime earner of foreign exchange followed by ourism. El-Sakka (2004), found ha Jordan is he hird larges remiance receiving counry in he Middle Eas afer Egyp and Morocco. I should be menioned, however, ha hese figures represen only officially recorded remiances and hey do no include remiances in kind and unrecorded remiances. While sources of he firs are relaively easy o esimae hrough he use of exensive surveys of emigrans, he laer is much harder o esimae. The scale of unrecorded remiances is unknown. Seccombe (1984) esimaes he acual remiances flows o be abou 60% higher of officially recorded remiances because of he inflow of remiances hrough unofficial channels and hose in kind. Talfaha (1985) ses acual remiances o be wo o hree imes he officially recorded remiances. To illusrae he imporance of remiances o Jordan, able (1) presens several indicaors. The able shows ha remiances are essenial for Jordan s balance of paymens. Excep for he period of he second gulf crisis, one can observe a seady rend of increasing remiance flows. Talfaha (1985), menions four facors responsible for he large increase in remiances: he increasing demand for Jordanian labor, huge developmen plans and rising wages in he hos counries, economic and poliical sabiliy in Jordan and sabiliy of he Jordanian Currency. Saring from 1987, remiances had declined reaching heir minimum in 1990 when Iraq invaded Kuwai. Level of 5

6 remiances had recovered quickly in Since hen, remiances has been growing reaching almos $ 2 billions in Year Remiances ($ millions) Remiances/ Expors (%) Table (1) Remiances in Jordan Remiances/ Impors (%) Remiances/ Consumpion (%) remiances/ GDP (%) Remiances Per-Capia ($) GDP Per- Capia ($) n.a. n.a n.a. Sources: Remiaces: IMF Balance of Paymens Saisics Yearbook Various issues Oher daa: IMF Inernaional Financial Saisics Various issues 6

7 For a long period of ime, remiances exceeded oal proceeds of Jordan s expors of goods. In 1977 remiances represened almos four imes he proceeds of expors. Currenly remiances represen abou 70% of Jordan s expors. Abou 45% of impors are recenly covered by he inflow of emigran remiances. The raio of remiances o privae consumpion is a measure of he exen o which remiances can be used o finance privae consumpion of domesic inhabians. Recenly remiances represen more han 35% of privae consumpion in Jordan. Compared wih GDP, remiances reached as high as 49% of GDP in 1977, and currenly represen abou 20% of GDP. According o El-Sakka (2004), hese raios are he highes by world sandards. Finally, he remiance per-capia indicaor shows how many US $ are remied for each inhabian in Jordan. To reflec growh in real erms, remiances mus increase a annual raes exceeding boh populaion growh and inflaion raes. El-Sakka (2004) found ha remiances per-capia in Jordan are he highes among he labor exporers in he Middle Eas. Daa in able (1) also show ha abou 20% of GDP per-capia is generaed from he inflow emigran remiances. The previous analysis shows he imporance of remiances o Jordan s economy. Appropriae policies o encourage emigrans o remi heir savings, in general, and use official channels of remiances in paricular, are hence essenial for he counry s macroeconomic sabiliy. The res of he paper aims a invesigaing he macroeconomic deerminans of remiances and possible impac of macroeconomic policy acions on heir inflows. The Model and Theoreical Consideraions Two groups of facors affecing he inflow of remiances o he counries of origin can be disinguished; micro-facors and macro-facors. The majoriy of he lieraure on remiances deal wih he micro facors. Micro approaches use a wide variey of variables o explain he flow of remiances including gender, age, marial saus, wage levels, per capia consumpion, number of households lef a home, lengh of 7

8 say abroad, educaional aainmen of poenial remiers, skill levels, occupaion ec. (See e.g., Lucas and Sark (1985), Roriguez, (1996), Izigsohn, (1995), Merkle and Zimmermann (1992)). Macro approaches aemp o idenify eiher he impac of remiances on macroeconomic variables in he counries of origin or he impac of macroeconomic variables on he inflow of remiances. Talafha (1985) uses he Keynesian income deerminaion model o esimae he effec of worker remiances consumpion, invesmen and impors. Using relevan mulipliers his resuls show ha remiances have significan muliplier impacs on hese variables. Burney (1987) aemps o analyze he effec of remiances on savings in Pakisan. Kandil, and Mewally (1990) address he impac of emigran remiances on he Egypian economy in he conex of he sandard Keynesian macroeconomic model. They aim a idenifying he facors deermining he magniude of he remiances muliplier and he impac of remiances on he major componens of he aggregae demand in Egyp. Looney (1990) examines he effec of remiances on macroeconomic developmen paerns in he Arab world. Glyoso (1993) examines he muliplied income effecs of remiances on macro variables in Greece. Gedeshi (2002), uses surveys of emigrans o examine he role of emigraion in he Albanian economy and heir moivaions for sending remiances back home. Lianos (1997) examines he flow of remiances and he facors deermining heir flow o Greece. In a sudy using daa from Algeria, Morocco, Porugal, Tunisia, Turkey, and Yugoslavia, Elbadawi and Rocha (1992) found he main deerminans of remiance inflows o be he level of income in he hos counry, he black marke exchange rae premium, he domesic inflaion rae, and he lengh of say abroad. The auhors did no find any evidence o suppor claims ha levels of income in he counry of origin or ineres rae differenials affec remiance inflows. In a panel sudy of remiances from Arab emigrans, El-Sakka (1998) found ha remiances were posiively relaed o economic growh in he hos counries and inflaion in he home counry. I was also found ha exchange rae differenials beween official 8

9 and black markes have a negaive impac on he inflow of remiances hrough official channels. El-Sakka and McNabb (1999), esimaed a macro model for oal inflow of remiances hrough official channels in Egyp. They found ha levels of income in boh hos and home counries have a posiive impac on he inflow of remiance o he home couny. They also found ha remiance flows are highly responsive o black marke premiums. The resuls also suppor he idea ha ineres differenials a home and abroad have a negaive impac on he inflow of remiances hrough official channels. They also find ha impors financed by remiances have higher income elasiciy and a relaively lower price elasiciy as compared o oher impors. The decision o remi is a fairly complicaed one. A ypical emigran has several opions; he has o decide on wheher o remi his savings o his home counry, or keep hem a he hos counry, or perhaps o remi hem o a hird counry, if here is a chance o do so. If he decided o remi o his home counry, he may also have o decide on wheher o remi his savings cash or in kind. If he decided o remi cash, he has o decide wheher o use official channels of ransfer (he banking sysem), or o use oher unofficial channels, including black markes for foreign exchange. A complemenary decision he remier has o make is abou he use of his remiances, i.e., wheher o use his savings for consumpion or invesmen. I is clear ha here are several facors ha affec all hese decisions including economic, social, culural and even poliical condiions. The model we propose o use for Jordan begins by assuming ha a migran has a arge level of remiances o his home counry R *. This arge depends on wo ses of facors, F 1 ; he non-policy variables se relaed o boh hos and home counries and F 2 ; he policy variables se, such ha; * R = f F, F ) (1) ( 1 2 9

10 F 1 se could include income levels, ineres raes and inflaion in boh hos counries and Jordan. Income levels in hos counries are expeced o have posiive impacs on he inflows of remiances, while ineres and inflaion raes are expeced o have negaive impacs on he inflows of remiances o he home counry. On he oher hand, he level of income in he home counry could have eiher a posiive or a negaive impac on he inflows of remiances, depending on he purpose of remiances. If remiances are mainly ransferred for family suppor, hen income levels in he home counry could have a negaive impac on he inflows of remiances, while inflaion would have a posiive impac. If remiances are, however, for invesmen purposes, we would expec a posiive relaionship beween levels of income and inflows of remiances, and a negaive impac of inflaion in he home counry. The mos relevan policy variables ha could influence he inflow of remiances o he home counry include a leas he following hree policy policies: - Exchange rae misalignmen, defined as deviaions from purchasing power pariy, which is assumed o negaively affec he inflow of remiances. As he degree of misalignmen increases, migrans will remi less and adjus heir arges. Since remiances are exchanged ino domesic currency, migrans may hold heir savings in foreign currencies or ime heir remiances wih exchange rae correcions. - Ineres rae differenial, which are defined as he difference beween foreign and domesic ineres raes. This is expeced o negaively impac he inflow of remiances, since low domesic ineres raes compared wih ineres raes in he hos or oher counries encourage remiers o wihhold heir savings in counries where ineres raes are higher. - General macroeconomic policies which include o Inconsisen moneary policy, here defined as he money supply over rend GDP. This could resul in excess supply of money and creaes inflaion. 10

11 Excessive money supply will no only mean more inflaion bu i will also indicae a policy failure, which may discourage remiances. o Inconsisen fiscal policy which could resul in a coninuous budge defici and all he risks associaed wih i. This sudy assumes ha emigrans monior macroeconomic policies in heir home counries and respond o hem by adjusing heir remiance arge levels. Because of he spread of Jordanian emigrans over he enire Gulf area, i is difficul o find a good proxy for variables relaed o he hos counries wihou some coss in erms of resul bias. For his reason, hose variables relaed o hos counries will be ignored. Inflaion in Jordan also was no included in he model due o he assumpion ha he effec of inflaion will be capured by money supply growh. Based on he analysis presened above, arge remiances is modeled as a log linear funcion of he following macro facors wih expeced signs shown above he differen variables; ( ± ) ( - ) ( - ) ( - ) ( - ) = θ & log B + ε (2) * log R a0 + a1 log y + a2 log + a3 log ( iw i ) + a4 log m + a5 where y = real income in he home counry θ = exchange rae misalignmen ( i w, i ) = hos counries and home ineres raes, respecively m& = growh of money supply over rend GDP as a proxy for moneary policy B = budge defici as a proxy for fiscal policy discipline. ε = an error erm I should be ha he model is buil on he assumpion ha each migran has his own arge remiances level R * in real erms. Since remied savings will be kep a 11

12 home, he sudy assumes ha his arge is denominaed in domesic currency, 3 e.g., suppose ha he arge of a migran is 1000 Jordanian Dinars (JD), he migran will coninue o remi unil he arge (1000 JD) is achieved. If during he curren period here is a change in he variables affecing he couner value in domesic currency, he real value of he arge may be differen from he acual (nominal) value of he achieved arge. In his case, a migran is assumed o ake some acions nex period o adjus for hese differences. For his reason he sudy assumes ha acual levels of remiances R are adjused o heir arge levels R * according o he following parial adjusmen scheme: * R = λ R R ) (3) r ( 1 where λ is he coefficien of adjusmen which akes a value beween 0 and 1 and a mean value equals ( λ / 1 λ ). The mean value of adjusmen coefficien reflecs he speed of adjusmen. I is assumed ha emigrans will parially reac o he deviaions beween heir acual and desired arges. If las year s acual remiances are higher han argeed levels, migrans will reduce heir remiances for his year and vice versa. Therefore, i is expeced ha here would be a negaive relaionship beween his year s remiances and lagged levels, since more remiances over his year s arge would mean less ransfer on he nex year and vice versa. By subsiuing (2) ino (3) we ge he following remiances equaion wih lagged remiances being among he regressors: log R = β + β log y + β logθ + β log( i 0 + β log B + β log R ε 3 w i ) + β log m& 4 (4) 3 I does no mean however, ha all remiances will be kep in domesic currency. In counries where deposis of foreign currency denominaions are allowed, under inflaionary condiions emigrans may prefer o keep heir savings in foreign currency. 12

13 The inclusion of lagged remiances could be defended on he following grounds: inconsisency of policy, lack of informaion and ineria may no allow migrans o remi he desired levels of remiances. In many cases changes in policies affecing remiances may occur wihou a noice. The flow of informaion o he hos counry may no allow migrans o correcly asses he effec of changes in policies on heir level of remiance arges, and hus here could be differences beween desired and acual remiances of migrans. Esimaion Resuls I is now well known ha performing a regression wih non-saionary series leads o a spurious regression. In many cases series need o be ransformed o induce saionariy; differencing is one mehod, and removing a deerminisic rend is anoher. 4 While he spurious regression is a serious issue, he pracice of differencing inegraed series o achieve saionariy and reaing he resuling series as he proper objecs of economeric analysis is no wihou coss. Prior o esimaion, a es of uni roos is applied o he series o check for saionariy of he variables. 5 If a variable is found o be non-saionary, a sequence of differencing is applied o he variable unil saionariy is achieved. We used Dicky Fuller, Augmened Dickey Fuller and Phillips-Parron ess of uni roo. Resuls are shown in able (2). The null hypohesis esed is ha he variable under invesigaion has a uni roo agains he alernaive ha i does no. The hree ess show ha mos of he variables are non-saionary in levels. The null hypohesis ha each variable has a uni roo could no be rejeced. Taking he firs difference of he variables renders all he variables o be saionary by all he hree ess as shown in able (2). Hence he variables are inegraed of order 1; I(1). Therefore, esimaion of equaion (4) should be based on he firs difference of he variables. Since he daa appear o be saionary in firs differences, no furher ess 4 Derending can eiher be made by including a funcion of ime as a regressor or by subracing a funcion of ime from all series used. (Banerjee e al (1993). 5 A series is said o be saionary if is mean, variance, and covariance are all unchangeable wih respec o ime. 13

14 are performed. We, herefore, mainain he null hypohesis ha each variable is inegraed of order one. In addiion, if hese variables have uni roos, hen we can exploi he idea ha here may exis a co-movemen in heir behavior and possibiliies ha hey will rend ogeher owards a long-run equilibrium sae. Having esablished he degree of inegraion of he variables, we proceed o es for coinegraion among he variables. Coinegraion is a es for equilibrium beween non-saionary ime series ha are inegraed a he same order. The concep of coinegraoin allows us o describe he exisence of an equilibrium, or saionary relaionship among wo or more ime series each of which is individually non-saionary. Table (2) Tes Resuls for Uni Roos Levels Variable Dicky-Fuller Augmened Dicky-Fuller Phillips-Parron R y θ m& ( w i i ) B Firs difference Variable Dicky-Fuller Augmened Dicky-Fuller Phillips-Parron R y θ m& ( w i i ) B The 5% significance level is The es for he null hypohesis ha here are r coinegraing vecors can be based on he maximum likelihood procedure proposed by Johansen (1998). We applied he 14

15 mehodology advanced by Johansen (1988) and Johansen and Juselus (1991, 1994). The Johansen-Jisulius procedure ess he exisence of more han one coinegraing vecor. The exisence of many coinegraing vecors may indicae ha he sysem is saionary in more han one direcion and hus more sable. Dickey e al (1994) argue ha he more coinegraing vecors here are, he more sable is he sysem; and i is desirable of he sysem o be saionary in as many direcions as possible. Since he Johansen s coinegraion procedure is sensiive o he choice of he lag lengh, he appropriae lag lengh needs o be specified. Using Akaike s final predicion error and Schawarz crierion, he appropriae lag lengh was found o be equal o wo. The maximum eigenvalue saisics λ max are used o deermine he number of coinegraing vecors r. The null hypohesis esed is ha here could be r coinegraing vecors among he variables of he model. Resuls of he Johansen-Jisulius procedure for deermining he number of coinegraing vecors are presened in able (3). 6 maximum eigenvalues, he race saisics, and he 90% criical values and he corresponding eigenvalues are also presened in able (3). According o he resuls, he hypohesis ha here are 4 uni roos can be rejeced in favor of 3 uni roos and hence 3 coinegraion vecors a he 10% level using boh maximum eigenvalue and race saisics. The hypohesis ha here is a leas one uni roo can no be rejeced agains he mainained hypohesis of no uni roos. A coinegraion rank r = 4 is hus seleced. The resuls confirm he exisence of a sable, long-run equilibrium relaionship. The Having esablished he coinegraing vecors for variables of he model we proceed o esimae he model using OLS. Resuls are presened in able (4). Looking firs 2 a he diagnosic saisics, he repored coefficien of deerminaion ( R ) is reasonably high. Mos of he variaions in remiances are explained by he variables included in he model. The F-saisic is significan a 1% level which indicaes ha he regression as a 6 The analysis of coinegraion is performed using he CATS Program by Hansen, H. and Juselius, K. (2002) in RATS 15

16 whole is significan. Given ha he Durbin-Wason es is biased owards finding no serial correlaion when he model conains a lagged dependen variable, he alernaive Durbin (h) saisic is calculaed. Resuls indicae ha he esimaed model is free from serial correlaion. Table (3) Johansen-Juselius Coinegraion rank analysis I(1) 7 Trace Saisic Maximum Eigenvalue λ max Eigen- value H 0 H 1 Trace Criical H 0 H 1 Criical max values 90% values 90% r = 0 r r = 0 r = r 1 r r 1 r = r 2 r r 2 r = r 3 r r 3 r = r 4 r r 4 r = r 5 r r 5 r = λ Table (4) Esimaion Resuls Variable Coefficien T-Saisic Diagnosic Saisics 7 Tess of he hypohesis ha here are a mos r coinegraing vecors 0 r <n and hus n-r uni roos, are based on: η = T r n i= r+ 1 log( 1 λ ), i r = 0,1,2,,n-0,n-1. This is he race saisic which is derived under he hypohesis ha here are r coinegraing vecors. The eigenvalues λ i are given by solving he λ eigenvalue problem:, where are he second momen marices of residuals and heir cross producs. Skk Sk 0k 1 0 S00 S = 0 S ij 16

17 Consan ** SEE = y ** SSR = y ** 2 1 R = θ ** F = 6.181** ( iw i ) ** DW = m& ** D(h) = B * R * 1 ** = significan a he 1% level. * = significan a h e 5% level. Since remiances o he home counry can depend no only on curren income a home, bu also on las year s income levels, a lagged income variable is included in he lis of explanaory variables. Concerning he level of income in Jordan, he resuls show ha here is a significan posiive relaionship beween he level of income a home and he inflow of remiances. As income levels in Jordan rise, he inflow of remiances increases, and vise versa. This indicaes, as menioned earlier, ha he majoriy of remiance flows o Jordan are for invesmen and no for family suppor purposes. 8 Rising income levels in he home counry reflec rising economic aciviy levels and hence, higher raes of reurn on invesmens a home. Under condiions of booming economic aciviy in he counries of origin, emigrans will send more remiances for invesmen purposes. This indicaes ha economic growh is a very imporan deerminan of he inflow of remiances o he counries of origin. Though i has been claimed ha differen incenives in erms of exchange and ineres raes, as well as oher incenives, are necessary for he inflow of remiances o he counries of origin, susained economic growh will no only mobilize domesic resources o invesmen bu also will mobilize differen forms of foreign capial including remiances. This resul is of grea imporance o he labor sending counries. As he rae of growh increases, he domesic resource gap will be narrowed as more emigrans will be willing o officially ransfer more savings for invesmen. The benefis of growh will be more obvious for counries which have a subsanial sock of emigrans abroad. 8 Similar resuls were reached by El-Sakka, M. & McNabb R. (1999) for Egyp. 17

18 Deviaions of exchange raes from heir real levels, proxied by purchasing power pariy, which reflecs he degree of misalignmen, have a significan negaive relaionship wih he inflow of remiances. The more he deviaion of exchange raes from heir purchasing power pariy levels, he less he inflow of recorded remiances o he counry of origin, and he more he inflow of remiances o informal channels will be. This resul is consisen wih evidence reached by previous sudies abou he effec of exchange raes on he inflow of remiances. Previous empirical sudies abou remiances show ha emigrans are very sensiive o exchange rae overvaluaion, e.g., in a sudy abou Egyp, El-Sakka (1999) found ha exchange rae overvaluaion, measured as he difference beween official and black marke raes, is very criical o he decision by emigrans o remi funds hrough official channels. Resuls indicae ha here is a significan negaive relaionship beween differences in ineres raes a home and abroad and he inflow of remiances. The greaer he differenials beween domesic and foreign ineres raes, he less likely ha emigrans remi heir savings. This reflecs he decisions aken by emigrans o maximize he rae of reurn on heir financial asse porfolio. If he rae of ineres is higher ouside Jordan, emigrans would be likely o keep heir savings abroad unil he ineres raes in he home counry are adjused o be in line wih foreign raes of ineres. Glysos (1996) claims ha arge emigrans, hough being risk seekers when considering he decision o work abroad, hey are risk averers when considering decisions o inves heir savings. Tha is why hey reain mos of heir savings abroad and only remi he necessary amoun needed for heir families mainenance. The mos ineresing resuls of he sudy are hose relaed o money supply growh over GDP rend and he budge defici. Boh coefficiens are found o be negaive. This im plies ha money supply growh and budge defici have a negaive impac on he inflow of remiances o Jordan. If money supply exceeds rend GDP levels, disequilibrium in he money and goods marke will occur creaing hus inflaionary 18

19 pressures. Inflaion creaes an unaracive environmen for all forms of foreign capial, including remiances. Likewise, coninuous budge defici are found o have an imporan direc effec on inflaion. The public finance approach o inflaion assumes ha he roos of inflaion can be found in large fiscal imbalances. Insufficien revenue collecion and limied access o domesic or inernaional financial markes end o increase governmens reliance on prining money as a primary source o finance he budge defici. This will cause prices o increase. Moreover, i has been argued ha he adverse effec of inflaion on he real value of fiscal defici reinforces he link beween fiscal deficis and inflaion. Coninuous budge defici is hus complicaing he problem of price sabiliy. Large budge deficis mean ha more inflaionary pressures are fed by he defici which in urn pus more pressures on he defici due o he needs for more funds o mainain he real value of public expendiures which, in urn causes defici o grow and hence money supply and so on. Conrolling he budge defici is hus, very imporan for inflaion and remiance inflows. These resuls show ha emigrans are sensiive o macroeconomic policy failures. Targe remiance levels of emigrans are adjused o macroeconomic policy acions. Consisen macro economic policies are hus very essenial for he inflow of remiances o labor sending counries. Finally, he coefficien of lagged remiances is found o be negaive as expeced. The mean adjusmen lag is found o be 2.3 years. This means ha acual remiances will be fully adjused o arge remiances in 2.3 years. This seems be a reasonable ime period given he srucure of LDCs and he qualiy of informaion available o emigrans. C onclusions and Policy Recommendaions 19

20 The aim of his sudy is o examine he impac of macroeconomic policies on he inflow of emigran remiances o Jordan. Differen macroeconomic policy failures are modeled in his sudy by exchange rae misalignmen, defined as he deviaion from equilibrium long run purchasing power pariy, ineres rae differenials, inconsisen moneary policy, defined as he growh of money supply over rend GDP levels, and inconsisen fiscal policy proxied by budge defici. The sudy found ha Jordanian emigrans are sensiive o macroeconomic policy. Targe remiance levels are negaively affeced by inconsisen macroeconomic policies. The following conclusions and policy recommendaions are drawn from he resuls above: Economic growh in he home counry of emigrans is an imporan deerminan of he inflow of remiances. Economic growh will help he counry arac differen ypes of capial and remiances. This in urn, helps o ease foreign exchange bolenecks and improve he posiion of he balance of paymens. Ineres raes policy should be carefully designed o arac remiances o official channels, policy makers should no only look a nominal ineres rae differenials, bu also nominal ineres raes should be adjused o reflec inflaionary pressures. Emigrans seem o be sensiive o exchange rae misalignmen. Policy makers need o be careful abou deviaions of exchange rae levels from heir equilibrium long run levels. Macroeconomic policy variables are found o be very imporan deerminans of inflows of remiances hrough official channels in Jordan. 20

21 References Awarani, H. and Kleiman, E. (1997) Economic ineracions among paricipans in he Middle Eas peace process The Middle Eas Journal; Washingon; Spring 1997; Vol. 51, issue 2, pp Burney, N. (1987) Workers Remiances from he Middle Eas and heir Effec on Pakisan s Economy, The Pakisan Developmen Review, Vol. 26, No. 4, PP Elbadawi, I. and R. Rocha Deerminans of Expariae Workers Remiances in Norh Africa and Europe. Working Paper No Counry Economics Deparmen, World Bank, Washingon, D.C. El-Sakka, M.I.T. (2004) Following he Flow of Remiances in he Middle Eas Chaper 13 in Wilson S. R. (ed.), Beyond Small Change: Making Migrans Remiances Coun, Iner-American Developmen Bank (forhcoming) (1998) Deerminans of Arab Migran Remiances. Journal of Social Sciences Vol. 26 No.3 pp El-Sakka, M.I.T. and R. McNabb (1999) The Macroeconomic Deerminans of Emigran Remiances, World Developmen, Vol. 27, no. 8 pp Gammelof (2002) Inernaional Migraion Remiances and Oher Financial Flows o Developing Counries, Vol. 40, No. 5, PP Gedishi, I (2002) Role of Reimances from Albanian emigrans and heir influence in he counry s Economy Easern European Economics, Vol. 40, no. 5, pp Glyoso, N. P. (1993) Measuring he Income Effecs of Migran Remiances: A Mehodological Approach Applied o Greece, Economic Developmen and Culural Change, Vol. No. pp Glysos, N. (2002) The Role of Migran Remiances in Developmen: Evidence from Medierranean Counries Inernaional Migraion Vol. 40, No. 1 pp IMF 2004 Counry Repor No.04/122: Jordan: 2004 Aricle IV Consulaion and Second Review under he Sand-By Arrangemen Saff Repor; Saff Saemen; Public Informaion Noice and Press Release on he Execuive Board Discussion; and Saemen by he Execuive Direcor for Jordan. Izigsohn J. (1995) Migran Remiances, Labor Markes, and Household Sraegies: A Comparaive Analysis of Low Income Household Sraegies in he Caribbean Basin, Social Forces, Vol. 74, No. 2, PP Kandil, M. and Mewally, M. F. (1990) The Impac of Migrans Remiances on he Egypian Economy Inernaional Migraion, Vol. 28, No. 2, PP Keely, C.B and Sake, B. (1984) Jordanian Migran Workers in he Arab Region: A Case Sudy of Consequences for Labor Supplying Counries he Middle Eas Journal, Vol. 38, no.4 pp Knowles, J. C. and Anker,. R. (1981) An Analysis of Income Transfers in a developing Counry: he Case of Keynia, Journal of Developmen Economics, Vol. 8, pp Lianos, P. T. Facors Deermining Migran Remiance: The Case of Greece Inernaional Migraion Review, Vol. 31, No. 1, pp Looney, R. L. (1989) Paerns of Remiances and Labor Migraion in he Arab World Inernaional Migraion, Vol. 27, No. 4, pp

22 Looney, R. L. (1990) Macroeconomic Impacs of Worker Remiances on Arab- World Labor Exporing Counries, Inernaional Migraion, Vol. 28, No. 1, pp Merkle L. and Zimmermann K. F. (1992) Savings, Remiances and Reurn Migraion, Economics Leers, Vol. 38, No. 1, PP Nisha, M. and Bilgrami, N. (1991) The Impac of Migran Workers Remiances on Pakisan Economy, Pakisan Economic and Social Review, Vol. 29, No. 1, PP Rodriguez, E. R. (1996) Inernaional Migrans Remiances in he Philippines, Canadian Journal of Economics, Vol. 29, April, PP Russell S. T. (1986) Remiances from Inernaional Migraion: A Review in Perspecive, World Developmen, Vol. 14, No. 6, pp Seccombe, I. A Jordan, (World Bibliographical Series, Vol. 55), Clio Press, Oxford, England. Talafha, H. (1985) The Effecs of Workers Remiances on he Jordanian Economy, METU Sudies in Developmen Vol. 12, No. 1-2, pp Taylor J. E. (1999) The New Economics of Labour Migraion and he Role of Remiances in he Migraion process, Inernaional Migraion, Vol. 37, No. 1, pp , Van Hear, Nicholas (1995) The impac of he involunary mass 'reurn' o Jordan in he wake of he Gulf Crisis, The Inernaional Migraion Review; Summer 1vol 29, issue 2, pp