Regionally targeted migration policy as an instrument for regional development: a general equilibrium assessment

Transcription

1 Regionally argeed migraion policy as an insrumen for regional developmen: a general equilibrium assessmen by N. H. Tran, E.L. Roos, J.A. Giesecke and J.R. Madden Cenre of Policy Sudies, Vicoria Universiy, Ausralia 1 Inroducion Regional policy-makers have a considerable number of policy insrumens wih which o influence he spaial disribuion of economic aciviy (see Figure 9.1 of Armsrong and Taylor, 2000). While many of hese policies are designed wih a regional policy objecive in mind, some policy opions involve he use of naional economic policies in a regional discriminaing way. In his paper we examine a policy ha falls ino he laer caegory, namely a regional discriminaing immigraion policy. The promoion of economic developmen in areas ouside of Ausralia s capial ciies has long been a policy ineres of he Ausralian Federal Governmen. The Ausralian Deparmen of Infrasrucure and Transpor (2013, p. 9) noes ha: Aside from ciy saes like Singapore and Monaco, Ausralia is he mos urbanised naion on earh. This issue increased in prominence in he 1990s when concerns were raised ha he gains from compeiion policy reforms were concenraed in major urban areas, paricular along Ausralia s eas coas (Indusry Commission 1995; O Conner e al. 2001; O Neill and McGuirk 2002). Since 1996/97, immigraion policy has been an explici insrumen in achieving he aim of regional developmen ouside Ausralia s sae capials (Cameron, 2011; Hugo, 2008). In paricular, a number of immigraion programs are ailored o encourage immigrans o sele in areas ouside of Ausralia s major populaion cenres. To encourage immigrans o sele in non-capial ciy regions, a number of migraion schemes have been designed o ease he graning of visas o migrans inending o sele ouside of Ausralia s major populaion cenres. These include boh a permanen skilled program and a emporary long-say business visa program (or T457). While hese programs are essenially designed o increase he Ausralian labour force s skill base, hey carry condiions of employmen which allows heir use for a second purpose of direcing skilled migrans o nonmeropolian regions. 1

2 T457 is a program which allows employers o bringing in skilled workers from overseas for up o four years o fill vacancies for which hey canno fill wih local residens. In principle, employers in any locaion can apply o his program, and hence i is no specifically designed o direc migrans o sele in non-meropolian areas. However, i can be, and is increasingly, used for ha purpose, because of wo main reasons. Firs, he T457 program imposes sricer condiions on labour marke mobiliy of visa holders. Primary 457 visa holders mus work in he same occupaion for he same sponsor as specified in heir visa nominaion, and mus no cease employmen for more han 28 consecuive days. Any change in employer mus be approved by he Ausralian Deparmen of Immigraion and Ciizenship (see DIAC, 2012a). Thus, T457 visa holders nominaed by a regional employer will have o work in he employer s region. Second, he program is flexible and can be adjus according o changes in labour demand. 1 Due o hese feaures and oher benefis o he economy, 2 he T-457 program is becoming increasingly imporan in Ausralia s migraion policy. In he seven years from 2004/05, he size of Ausralia s annual T457 visa inake grew almos hreefold, rising from 17 housand o more han 48 housand. The program makes up 47.3% of gross skilled migran inake in he financial year 2011/12, and is projeced o make up jus under 60% of he gross skilled migran inake for he period up o 2021/22 (DIAC, 2012f). While by design he T457 program can help address shor-erm skill shorages, heir efficacy in promoing regional developmen is no well explored. Sudies on immigraion and regional developmen have been mosly descripive or qualiaive in naure. Surveys, inerviews or case sudies were used o explore demographic and labour marke characerisics of migrans; reasons for migraion; facors ha influence migrans o sele in regional areas; migrans mobiliy; level of migrans and heir employers saisfacion; and he social aspecs of migraion. 3 There have been few sudies focusing specifically on he T457 program. Khoo e al. (2005), Khoo and McDonald (2006) is a survey on reasons for migraion, employmen circumsances and migraion oucomes of T457 visa holders. Access Economics (2002a, 2002b) uses a macro-economeric model o assessed economic impacs of T457 program on he governmen budge and on Ausralia s living sandard, bu he sudy focuses on he 1 For example, according o DIAC (2012f), he level of T457 inakes declined from 44 housand in 2008/09 o jus over 26 housand during he economic downurn during 2009/ /11 due o he global financial crisis, and hen rose again o over 48 housand in 2011/12 when he economy recovered from he crisis. 2 See analysis of he impacs of he T457 program on governmen budge and Ausralia s living sandard in Access Economics (2002a, 2002b). 3 See, for example, a review of research on he use of skilled migraion o address regional skill shorages by Cameron (2011). 2

3 economy as a whole, no on regional developmen. The conribuions of migrans o regional Ausralia have been saed in erms of hem seling here on arrival (Cully 2010). I is believed ha Skilled migrans who sele in regional Ausralia help suppor and grow regional economic aciviy. They help susain and srenghen regional communiies by filling criical skills shorages. They add vialiy, diversiy and innovaion o our regional economies. (Crean, 2012, p.151). This seems especially so for T457 visa holders because, compared o he Ausralian average, hey end o be younger, have higher labour force paricipaion rae, lower unemploymen rae, and hey concenrae in higher-skilled professional and managerial occupaions (Access Economics, 2002b; ABS 2008a) However, i is no clear wheher regionally-argeed T457 programs can in any meaningful way impac on he spaial disribuion of regional economic aciviy. Firs, T457 visa holders have been found o be a highly mobile group. On average, abou 13% of hem emigrae each year (DIAC, 2012f). Of hose who say in Ausralia, Khoo and McDonald (2006) found ha over 7% change heir place of residence wihin one year. A panel daa of T457 visa holders for he period show ha on average 5.7% of hose saying in he counry move ou of non-capial regions each year (DIAC 2012g). 4 Second, here could be a displacemen effec in he labour marke, where emporary skill migrans ake up jobs which could have been filled by local residens. 5 This paper provides he firs quaniaive assessmen of he efficacy of emporary skill migraion program in promoing economic developmen of regions ouside of meropolian areas in Ausralia. We use TERM-M, a general equilibrium model wih exensive regional and migraion deails, o run 8 policy scenarios agains a common baseline forecas for he period 2010/ /22. In he firs seven policy scenarios, we simulae a once-off argeed migraion program, whereby he 2012/13 inake of 457 visa holders is increased by 1000 persons in each of he model s seven non-meropolian regions. In he eighh scenario we simulae he same 1000-person increase in 457 numbers, bu his ime wihou direcing hem ino any paricular region; raher, hey will sele across he model s 15 regions according o curren selemen paerns. Simulaion resuls show ha over ime, iner-regional migraion 4 This is only slighly lower han he mobiliy of permanen regional skilled migrans. Surveys by DIAC (2005a, 2005b) found ha around 6% and 8% of principal applicans of he Sae Designaed Migraion Scheme and Regional Sponsored Migraion Scheme respecively inend o move o anoher region in 12 monhs ime. 5 See, for example, ABC (2012) for recen high-profile cases where labour agreemens were used o bring in emporary 457 workers o work on resource projecs in regional Wesern Ausralia. See also Birrell and Heard (2012) for discussion on he displacemen effecs in he medical profession. 3

4 and labour marke displacemen effecs resul in a weakening of he increase in regional supply even in he shor run. Over a decade his diminishing of he regionally-arge program s effecs is compounded by around 60 per cen of emporary visa holders reurning overseas, so ha for he argeed region here is lile by way of a long erm regional legacy effec. The paper is organised as follows. Secion 2 describes he heory and he daabase of he TERM-M model. Secion 3 discusses he policy shocks implemened in he simulaions. Secion 4 provides a brief overview of he effecs of he shocks on all non-capial regions in simulaions. Secion 5 provides a deeper analysis for he resuls, aking he Res of Wesern Ausralia as an example. Secion 6 furher discusses macroeconomic effecs of he shocks on all regions. Secion 7 concludes he paper. 2 The TERM-M model 2.1 Theory Broadly, he research quesion requires a model ha conains boh regional economic deail and labour marke deail. Hence, we use a boom-up muli-regional model based on TERM (The Enormous Regional Model) (Horridge, 2011) inegraed wih a deailed labour supply module based on he labour supply heory pioneered by Dixon and Rimmer (2008). TERM explicily capures he behaviour of indusries, households, invesors, governmen and exporers a he regional level. The heoreical srucure of TERM follows he familiar neoclassical paern common o many applied general equilibrium models. Producers in each region are assumed o minimize producion coss subjec o a producion echnology ha allows subsiuion beween primary facors (labour, capial and land) and beween geographical sources of supply for specific inermediae inpus. Demand for effecive inpu of labour o each regional indusry is defined over labour disinguished by occupaion. A represenaive household in each region purchases goods in order o obain he opimal bundle in accordance wih is preferences and disposable income. Invesors seek o maximize heir rae of reurn, while demand by foreigners is modelled via expor demand funcions ha capure he responsiveness of foreigners o changes in expor supply prices. TERM s heoreical and daa srucures are well documened in Horridge e al. (2005) and Horridge (2011), and so we do no expand furher on he srucure of TERM in his paper. 4

5 Migraion deails are included in he labour supply side of he model. The labour supply heory imposes a sock/flow dynamic on highly disaggregaed labour marke groups. Beween years, sub-populaions, defined by age, region, labour force saus and visa saus, ransi beween age and visa saus caegories. Wihin years, hese same sub-populaions make uiliy-opimising offers o various labour marke aciviies, such as working in a paricular occupaion, or leaving he labour force, or go overseas. Five equaions are fundamenal o he operaion of he labour marke module. The firs equaion groups people a he sar of year o caegories based on heir labour marke aciviies o 6 in region r 7 in year -1, aking ino accoun changes in heir age groups a 8 and visa ype v 9 via he mediaion of a ransiion marix. 10 where 1 CAT( orva,,, ) ACT( orvvaa,,, ) * T ( vvaava,,, ) for all (o,r,v,a) o New (1) aa AGE vv VISA = CAT = exogenous for all (r,v,a) (2) ( New,,, r v a ) CAT ( orva,,, ) is he sock of working age populaion a he sar of year, grouped by heir labour marke aciviy o and region of residence r in year -1, and age and visa ypes a he sar of year. CAT( New,,, r v a ) is he new enrans o he working age populaion. The new enrans consis of people urning 15 in year, and new permanen and emporary migrans from overseas. Noe ha he New elemen is an addiion o he labour marke funcions (o) in equaion (1). 6 Labour marke aciviies consis of major occupaional groups, shor-erm unemploymen, long-erm unemploymen and no in he labour force. The major occupaional groups are: Managers; Professionals; Technicians and Trades Workers; Communiy and Personal Service Workers; Clerical and Adminisraive Workers; Sales Workers; Machinery Operaors and Drivers; and Labourers Ausralian regions (he ACT plus 7 capial ciies and 7 Res of he remaining 7 Ausralian saes and erriories). 8 Namely, seven10-year working age groups, saring from 15 o Namely, Ciizen; Skilled permanen visa; Family visa; Humaniarian visa; Oher permanen visas; Suden visa; Temporary long-say business visa 457; Oher emporary visas; and New Zealanders. 10 Noe ha, CCCCCC (NNNNNN,rr,vv,aa) is deermined exogenously. CAT( New,,, r v a ) is he new enrans o WAP. The new enrans consis of people urning 15 in year, and new permanen and emporary migrans from overseas. Noe ha he New elemen is an addiion o he labour marke funcions (o) covered by equaion (1). 5

6 1 ACT( o, r, vv, aa) is he number of people aged aa holding visa vv, who performed aciviy o in region r during year 1; 11 T ( vvaava,,, ) is he ransiion marix showing he probabiliy of someone age aa holding visa vv in year -1 o move o age group a wih visa v a he sar of year. This marix will be srongly diagonal. However, i will also conain imporan off-diagonal componens describing elemens of governmen policy. For example, every year some proporion of people in he emporary business long-say 457 caegory may apply o and be graned a permanen skilled visa. Noe ha equaion (1) does no deermine how people change heir labour marke saus or region of residence from year o year. These changes are deermined by an opimisaion problem, where a he beginning of year people in all caegories offer heir labour o a labour marke aciviy in a region o maximise heir uiliy. This resuls in a labour supply funcion in he form of equaion (2), which describes how people in caegories CAT ( oo, rr, v, a) offer heir labour o specific labour marke aciviies o in region r in year in response o differences in relaive incomes in hose labour marke aciviies. η ( B( oorrvaor,,,,, ) * ATW( ) ) or, L( oorrvaor,,,,, ) = CAT( oorrva,,, ) * η ( ( ) ( ) ) B oorrvaqp,,,,, * ATW qp, q p (2) where CAT is he sock of working age populaion in each domesic labour marke ( oo, rr, v, a ) caegory oo in domesic region rr a he sar of year. L( oorrvaor,,,,, ) is he labour offer from all labour marke caegories oo in region rr, o all labour marke aciviies o in region r. In addiion o offering heir labour o domesic aciviies, our model heory also allows people o emigrae. Tha is, in addiion o domesic labour marke funcions oo and domesic region rr in caegories CAT ( oo, rr, v, a ), he 11 A noe on our se index noaion: hroughou his secion we use double leers (such as oo, rr, vv) o denoe he iniial posiion, and he single leer (such as o, r, v) o denoe he final posiion of a caegory or aciviy. For example, in he coefficien T ( vvaava,,, ), vv and aa denoe he visa ype and age of a group of people during year -1, and v and a denoe heir visa and age a he beginning of year. 6

7 new aciviy (o) and region (r) in L ( oorrvaor,,,,, ) ) include emigraion o Res of he World. Noe also ha all caegories a he beginning of year can offer heir labour o all aciviies during year, excep: (i) employed people can say employed or become shor-erm unemployed, bu canno become long-erm unemployed; (ii) shor-erm unemployed people can become employed or long-erm unemployed, bu canno say shor-erm unemployed; (iii) new enrans can become employed or shor-erm unemployed, bu no long-erm unemployed; 12 and (iv) we assume ha immigrans do no emigrae in he firs year of heir arrival. ATW is he real afer-ax wage rae in aciviy o in region r. ( or, ) ( (,,,,, ) * ( qp, ) ) oorrvaqp q p B ATW is average real afer-ax wage for all caegories (oo,rr,v,a). η is a parameer ha reflecs he ease wih which aduls can shif beween aciviies. Following Dixon and Rimmer (2010), we se η a 2 for all domesic aciviies. I is assumed ha here is a higher level of difficuly for shifing beween domesic and overseas aciviies, and hence η is se a 1 for overseas aciviies. B( oorrvaor,,,,, ) is a occupaional and regional mobiliy marix which shows he probabiliy of people changing occupaion or region of residence even in he absence of changes in relaive prices. I capures exogenous non-wage facors, such as preferences, ha may moivae people from caegory (oo,rr,v,a) o offer heir labour o aciviy o in region r. Equaion (3) ranslaes labour supply by persons ino labour supply in hours.,,, (,,, orva) LH = HPP LS ( orva) (,,, orva) (3) where HPP is he number of hours worked per person in occupaion (o,r) by persons wih ( orva,,, ) visa saus v, age a. LH is he number of hours of labour supplied o occupaion o in region r by ( orva,,, ) persons wih visa saus v, age a. 12 The excepions (i) o (iii) sem from he fac ha caegories are grouped by heir labour marke aciviies in year -1, and by definiion, shor-erm unemploymen is unemploymen of no more han 12 monhs. 7

8 LS is he oal number people in aciviy o in region r by visa saus v and age a. ( orva,,, ) The disincion beween persons and hours is imporan because here is a big difference in he number of working hours per person per week beween visa ypes. For example, while he Ausralia-wide average number of working hours per person per week is approximaely 33 hours (ABS, 2010a), he number is abou 20 for Suden visa holders (ABS, 2011a). In equaion (4), he oal number of working hours supplied o occupaion o in domesic region r is deermined as: LS = ( or), LH, v VISA a AGE (,,, orva) (4) Finally, Equaion (5) links he labour supply module wih he TERM model. LD LS * FLAB or = (5) ( or, ) ( or, ) (, ) Equaion 5 equaes labour demand o labour supply and a shif variable. Labour demand is deermined via he facor demand equaions specified in he TERM model. Labour supply is deermined in he labour marke module via Equaion 4. Wih he shif variable se exogenously, labour demand is equal o labour supply and any disequilibrium beween labour demand and supply will cause he wage o adjus. TERM-M is he firs model o conain a deailed reamen of labour supply by visa ype and he ransiion beween visa ypes for Ausralia. The disincion beween visa ypes is imporan for he assessmen of he effecs of migraion policies on he economy for wo reasons. Firs, migran groups differ in heir labour marke characerisics. For example, migraion daa (ABS, 2011a; DIAC 2012f, 2012g) have shown ha compared wih oher migran ypes, skilled migrans ypically have higher labour marke paricipaion rae and lower unemploymen rae, and are more likely o work in managerial and professional jobs (ABS, 2011a). They also work more hours per year han holders of suden and working holiday visas. Second, holders of differen visa ypes have differen regional and occupaional mobiliy. By definiion, emporary migrans have a much higher emigraion rae han permanen residens. Temporary visas are ofen a pahway o permanen visas, and once a emporary visa holder becomes permanen residen, heir propensiy o emigrae falls. Domesic mobiliy also differs beween differen visa ypes because of differen visa condiions. For example, as discussed earlier, he 457 visa imposes more resricion on visa holders movemen beween occupaions and regions han oher visa ypes. Therefore, i is 8

9 imporan no only o disinguish hese visa ypes, bu also o model he changes among hem, which is he funcion of he visa ransiion marix. The model is solved wih GEMPACK a general equilibrium modelling package (Harrison and Pearson, 1996). 2.2 The model daabase The daabase for he TERM module of he model was buil using Ausralia s inpu-oupu (IO) able for he year 2005/06 (ABS, 2009). For he process of convering he IO daa o he forma suiable for he deailed regional daa for he TERM model, see Horridge (2011). The daabase is hen updaed o he year 2011 using available hisorical economic daa for he period 2006/ /11. The labour supply module (equaions 1 o 5) requires he following marices: he lagged aciviy marix ACT 1 ( o, r, vv, aa) ; he age and visa ransiion marix T ( vvaava,,, ) ; he marix of new enrans CAT New,,, r v a ; he occupaional and regional mobiliy marix B ( oorrvaor,,,,, ), and he hour-worked marix HPP. 13 The daabase consrucion process aims o firs creae all of ( orva,,, ) hese marices for he period 2007/08, and hen updae hem o he period 2010/11. The compilaion of hese marices is described below. 1 The compilaion of he lagged aciviy marix ACT( o, r, vv, aa) for he year 2006/07 sared from Census 2006 (ABS, 2006). The Census provided informaion on age, region of residence, labour force saus, occupaion of employmen for Ausralian ciizens and non-ciizens, bu no for deailed visa ypes. The non-ciizen par of he Census was hen disaggregaed ino he required 8 non-ciizen visa ypes using sock daa for emporary and permanen visa holders (DIAC, 2011) and confidenialised uni record files from he Characerisics of Recen Migrans surveys in 2007 and 2010 (ABS 2008a, 2010a). Daa for boh ciizens and nonciizens were hen scaled o ensure ha hey sum o ABS s working age populaion saisics for he financial year 2006/07 (ABS, 2012a-2012d). Nex, he ransiion marix T ( vvaava,,, ) was esimaed as a produc of he age ransiion marix and he visa ransiion marix. The age ransiion marix was esimaed using demographic daa on age disribuion and deah raes (ABS, 2011, 2010d). The visa ransiion marix was 13 The marix of average afer-ax wage rae ATW ( or, ) is deermined endogenously in he model. 9

10 esimaed wih wo seps. In he firs sep, he iniial ransiion marix was derived from informaion conained in ABS surveys of recen migrans for 2007 and 2010 (ABS, 2008a, 2010a) on visa ypes of visa holders on arrival and a survey ime, and DIAC panel daa for he period for over 317 housand emporary visa holders (DIAC, 2012h). In he second sep he iniial visa ransiion marix was adjused so ha i racked he sock and flow migraion daa supplied by DIAC for each of he years in he period (DIAC, 2011; 2012f, 2012g). The marix of new enrans CATNew,,, r v a was creaed using wo daa sources: (i) ABS s demographics daa (ABS, 2011) for he number of domesic new enrans, i.e. people 14-year of age, urning 15 in he following year; and (ii) DIAC daa on inflows and ouflows of migrans (DIAC, 2012f). The occupaional and regional mobiliy marix B ( oorrvaor,,,,, ) was esimaed using labour mobiliy daa for 2008 and 2009 (ABS, 2008b; 2010b) and Census 2006 informaion on people s regions of residence on Census nigh and one year earlier. The average working hours per person per year marix HPP was creaed using ( orva,,, ) informaion from CoRM 2010 (ABS, 2010a, 2010c) and Tan e al. (2009). Up o his poin he compilaion of he labour supply daa base for he year 2007/08 was 1 compleed. The ime dependen marices ACT( o, r, vv, aa), N,, rva and HPP ( orva,,, ) in he daabase were hen updaed o he period 2010/11 in order o mach he updaed TERM daabase, using labour marke, demographic and migraion saisics for he period (ABS, 2012a- 2012d, 2011; DIAC, 2011, 2012g). 3 Simulaion design To invesigae he efficacy of he T457 emporary visa programs as an insrumen of regional developmen, eigh simulaions are underaken. In he firs seven simulaions a argeed migraion program is simulaed, whereby he increase in T457 visa inake is direced o each of he model s seven non-meropolian regions. In erms of equaion (3) hese policy shocks 2012/ 13 represen a 1000 person expansion in he value of CAT ( New,, r T 457, a ), where r is he region lised in Table 1. In he discussions below hese policy scenarios are called argeed policies. In he eighh simulaion he same 1000-person increase in T457 numbers is simulaed, bu his 10

11 ime wihou direcing hem ino any paricular region. I is assume ha he new arrivals will sele across he model s 15 regions according o curren selemen paerns. This policy scenario is called unargeed policy. Consisen wih equaion (3), hese T457 visa holders hen supply labour o occupaions and non-employmen aciviies, boh wihin he region in which hey iniially arrive, and in oher regions. The percenage deviaions in he working age populaion (WAP) in he policy year 2012/13 due o he increase of 1000 persons in he argeed regions are repored in Table 1. Sim Table 1. Regional working age populaion under alernaive policy scenarios in 2012/13 Region Baseline level ( 000 persons) (a) Under he 7 argeed simulaions (increase of 1 hous. pers. in each region) % increase in WAP Under he unargeed simulaion (increase of 1 hous. pers. in he whole counry) Number of new arrivals o he region ( 000 persons) (b) % increase in WAP (1) (2) (3) (4) (5) (6) 1 Res of New Souh Wales (RoNSW) 2, Res of Vicoria (RoVIC) 1, Res of Queensland (RoQLD) 2, Res of Souh Ausralia (RoSA) Res of Wesern Ausralia (RoWA) Res of Tasmania (RoTAS) Res of Norhern Terriory (RoNT) Ausralia 18, Noes: (a) Model daabase, compiled based on daa sources discussed in Secion 2; (b) The number of new arrivals o each of he region is proporional o he share of he region in oal new T457 visa inake in he baseline case. For example, he number of housand new arrivals in RoNSW (row 1, column 5) corresponds o RoNWS s share of 3.78% in oal T457 inake in 2012/13. Noe ha he sum of arrivals o all non-capial ciy regions in his column is 0.17 housand person. The remaining 0.83 housand persons go o capial ciies. 4 Simulaion resuls Under all 8 policy scenarios, he oal number of working age populaion increases by 1 housand persons in he year 2012/13. However, he impacs on regional GDP are differen beween regions (Figure 1). This is because, as can be seen from column 4, Table 1, he same absolue addiion o WAP resuls in differen deviaions for regional WAP relaive o base case. This in urn, causes differen deviaions in regional employmen and GDP. 11

12 Figure 1. Regional GDP under alernaive simulaions (Percenage deviaion from baseline) Res of New Souh Wales RoNSW argeed policy Unargeed policy Res of Vicoria RoVIC argeed policy Unargeed policy Res of Queensland RoQLD argeed policy Unargeed policy Res of Souh Ausralia RoSA argeed policy Unargeed policy Res of Wesern Ausralia RoWA argeed policy Unargeed policy Res of Tasmania RoTAS argeed policy Unargeed policy 12

13 Res of Norhern Terriories RoWA argeed policy Unargeed policy In he shor run, argeed migraion programs cause larger posiive deviaions in regional real GDP han he unargeed program. However, in he long run, regional GDP resuls differ lile beween he argeed and unargeed programs. To uncover he drivers of hese resuls, in he secions below we will examine more closely he resuls in he argeed and unargeed migraion programs for RoWA. 14 The mechanisms of he impacs of he policies on RoWA are similar o hose in oher regions. 5 The effecs of T457 migraion program argeed a RoWA This secion discusses he resuls from he unargeed and RoWA argeed policy of increasing he inake of T457 visa holders by 1000 person. 5.1 Naional resuls Our ulimae aim is o undersand he consequences of he regionally argeed program for regional developmen. Neverheless, as we shall see, he naional macro consequences of he policies have a bearing on he regional economic consequences. While i is rue ha he region in which he 1000 persons expansion occurs has some influence on he naional macroeconomic resuls, in he broad boh scenarios can be viewed a he naional level as a 1000 person expansion in he working age populaion in he firs year of he policy (Figure 2). Thereafer, he WAP deviaions seadily decline. This reflecs he operaion of equaion (3). Each year, some proporion of T457 visa holders reurn overseas 15. In erms of equaion 14 In recen years, RoWA has been he region wih high inake of 457 visa holders o work on resource projecs (see, for example, ABC 2012). 15 Some also die, bu his explains only a small fracion of his decline. No only 457 visa holders. Via equaion (1) each year some proporion of 457 visa holders who remain in he counry ransiion o oher visa caegories. 13

14 (3), his fracion is represened by he percenage difference beween L and CAT 16. For T457 visa holders, he annual propensiy o reurn overseas is approximaely 13.5%. A he same ime, via equaion (1) he propensiy for T457 o ransiion o oher visa caegories is approximaely 18.6%. These visa caegories have lower emigraion propensiies han he T457 caegory. This accouns for he gradual sabilisaion of he naional WAP deviaion in Figure 2. The deviaion in employmen under boh policies closely follows he deviaion in WAP. Figure 2. Naion-wide working age populaion and employmen under alernaive policies (number of persons, deviaions from baseline) WAP, unargeed policy Employmen, unargeed policy WAP, RoWA argeed policy Employmen, RoWA argeed policy Figure 3 repors resuls for employmen, capial sock, real GDP and he average real wage a he naional level. These resuls are very similar under boh argeed and unargeed migraion policies. Common feaures Figure 3 for he wo policy scenarios are as follows. The posiive deviaion in working age populaion generaes a posiive deviaion in labour supply and employmen (Figure 3). In 2012/13 Ausralia s working age populaion is projeced o be 18,967 housand persons respecively (las row, Table 1). The increase in 1000 persons represens % increase in he WAP bu abou 0.009% for employmen, because T457 visa holders have a 16 Movemen in he relaive wage erm in equaion (2) exers only a small influence on he fracion. 14

15 higher labour force paricipaion rae han economy-wide average. 17 Similar o he deviaion in WAP, he employmen deviaion peaks in he firs year of he simulaion. Thereafer i gradually declines as he T457 visa holders emigrae. A proporion of T457 visa holders also become permanen migrans, who have lower emigraion raes han he emporary T457 visa holders, hence he increase in 2012/13 T457 inflows generaes a long-run permanen increase in employmen. Capial socks are sicky in he shor run, since hey can only deviae from baseline if invesmen deviaes from baseline. Wih he capial sock responding only gradually o he posiive employmen deviaion, in he shor run he labour/capial raio rises relaive o baseline (Figure 3). The shor-run increase in he labour/capial raio causes he marginal produc of capial, and wih i, he rae of reurn on capial, o rise relaive o baseline. Our model assumes a posiive relaionship beween invesmen and raes of reurn. Hence, invesmen rises relaive o baseline in he shor run. Over ime, he posiive deviaion in real invesmen causes he capial sock o rise relaive o baseline (Figure 3). By he end of he simulaion period he employmen and capial deviaions are of a similar magniude, signalling a reurn of he labour/capial raio o close o is baseline value. Wih boh employmen and he capial sock above baseline, he real GDP deviaion is posiive (Figure 3). However, we noe ha under he RoWA argeed policy he shor-run deviaion for wagebillweighed employmen and GDP is lower, and he long-run deviaion for capial sock is higher, han under he unargeed policy. This is because he negaive deviaion in he real wage due o he increase in labour supply is slighly larger in he RoWA argeed policy han under he unargeed policy (Figure 3). Indusries in RoWA are relaively more capial inensive han economy-wide average, and hence he region has relaively less elasic aggregae supply curve han economy-wide average. Under he RoWA argeed policy, he large negaive deviaion in he RoWA wage rae causes a larger negaive deviaion in he average naional real wage han ha under he unargeed policy. This explains he lower posiive deviaions in wagebill-weighed employmen and, in urn, a lower posiive deviaion in real GDP under he RoWA argeed policy han under he unargeed policy. However, as indusries in RoWA is relaively more capial inensive han economy-wide average, he RoWA argeed policy causes a larger long-run aggregae capial sock posiive deviaion 17 According o he Characerisics of Recen Migran Survey (ABS, 2008a, Table 2), he labour force paricipaion rae of 457 visa holders is 87.1%, whereas i is approximaely 66% for he whole working age populaion. 15

16 han ha in he unargeed policy. This explains he long-run convergence of real GDP under he wo policies. Figure Naion-wide real GDP, employmen, capial sock and average real wage under alernaive simulaions (Percenage deviaion from baseline) Employmen, RoWA argeed policy Capial sock, RoWA argeed policy Real GDP, RoWA argeed policy Average real wage, RoWA argeed policy Employmen, non-argeed policy Capial sock, non-argeed policy Real GDP, non-argeed policy Average real wage, non-argeed policy 5.2 Resuls for Res of Wesern Ausralia This secion examines he economic impacs of he argeed and unargeed migraion policies on he RoWA region Working age populaion and employmen Figure 4 repors he deviaion in WAP and employmen in RoWA under alernaive policy simulaions. As expeced, deviaions in WAP and employmen are much higher under he argeed program han under he unargeed one, because much fewer people arrive o he region under he laer (see Table 1). However, even in he firs year of he argeed policy wih 1000 person arrival o he region, he deviaions in boh WAP and employmen are less han he 1000 persons. Afer ha he posiive deviaions in boh WAP and employmen decline rapidly over ime, wih he deviaion in employmen says slighly higher han ha in WAP. 16

17 Figure 4. Working age populaion and employmen in RoWA under alernaive simulaions (Number of persons, deviaion from baseline) WAP in RoWA, non-argeed policy WAP in RoWA, RoWA argeed policy Employmen in RoWA, non-argeed policy Employmen in RoWA, RoWA argeed policy To undersand he forces behind he WAP deviaion in RoWA in he argeed migraion program, we decompose i ino he conribuions of he flows in and ou of he region (Figure 5). The inflows include new enrans ino WAP (boh domesic and foreign) and inerregional immigrans 18. The ouflows include emigraion o oher saes and overseas. In he firs year of he argeed policy, here is a posiive conribuion of new enrans ino he RoWA region compared wih he baseline due o he increased inflows of T457 visa holders ino he region. However, he conribuions from inerregional migraion are negaive because he inflows from oher saes decline and he ouflows o oher sae increase. From he second year of he policy onwards, overseas emigraion from he region also increases, conribuing negaively o he WAP deviaion in he region. The negaive conribuions from overseas emigraion coninue hroughou he simulaion period. The negaive conribuions from inerregional migraion coninue unil abou 2016, and hen reverse o become slighly posiive. 18 There are also deahs, bu he deviaion in he deah numbers is negligible, and hence is no included in Figure 5. 17

18 Figure 5. Conribuion of populaion flows o he deviaion of WAP in he Res of Wesern Ausralia under he RoWA argeed policy (Number of persons, deviaion from baseline) 1,200 1, New Inersae immigraion Inersae emigraion Overseas emigraion WAP The main reason for he deviaions in inerregional migraion is he deviaions in he real afer-ax wage rae in he region compared wih he wage rae in oher region. In he shor run, he increase in labour supply in RoWA compared wih he baseline causes a negaive deviaion in he region s wage raes (Figure 6). According o equaion (3), his discourages people o move from oher regions o RoWA, and encourages more people o move from RoWA o oher regions. As people move from RoWA o oher regions, he wage raes in oher regions also fall (Figure 6). The region wih he second larges wage fall is Perh, because Perh is he bigges desinaion for people moving ou of RoWA. 19 The wage raes in oher regions only fall slighly. In he long-run, as he number of WAP comes back o he baseline level, he wage raes also come back o he baseline level. 19 According o Census 2006, more han wo hird of hose moving ou of RoWA during he year before he Census nigh moved o Perh (ABS, 2006). 18

19 Figure Wage rae by region under he RoWA argeed policy (Percenage deviaion from baseline) Sydney RoNSW Melbourne RoVIC Brisbane RoQLD Adelaide RoSA Perh RoWA Hobar Employmen deviaion is posiive because he increase in labour supply causes he wage rae o fall and employmen o rise. However, employmen deviaion is slighly higher han ha of WAP because he newly-arrived T457 visa holders have higher han average labour force paricipaion rae and lower han average unemploymen rae Oher macroeconomic effecs Table 2 repors macro resuls for RoWA under argeed and unargeed policies. In he firs year of he policy he posiive deviaions in all macroeconomic indicaors for RoWA under he argeed policy are larger han under he unargeed policy. The main reason is ha employmen deviaion under he RoWA-argeed policy is much larger han ha under he unargeed policy. In he RoWA-argeed scenario, wagebill-weighed employmen deviaion is 0.25% (row 2, Table 2), slighly larger han he deviaion in employmen by person of 0.22% discussed in he previous secion. This is because T457 visa holders end o work in managerial and professional job wih higher han average wage raes. The posiive deviaion in employmen causes posiive deviaions in GDP and invesmen under he same 20 According o he Characerisics of Recen Migran Survey (ABS, 2008a, Table 2), he labour force paricipaion rae of 457 visa holders in November 2007 was 87.1%, whereas i was 66.2% for all visa caegories, including ciizens. As regard unemploymen rae, he unemploymen rae of 457 visa holders are lower han economy-wide average because design all primary 457 visa holders are employed, and visa holders comprise abou half of all 457 visa grans over he period 2007/ /12 (DIAC 2012i). In our baseline forecass for 2012/13 unemploymen rae of 457 visa holders is 3.4%, whereas he economy-wide average unemploymen rae is 5.1%. 19

20 mechanisms as discussed in Secion 5.1 for naional resuls. Privae and public consumpions rise wih GDP. All deviaions in invesmen and consumpion are lower han he deviaion in GDP, hence he balance of rade moves oward surplus. Impors rise o caer for increased demand from producion, invesmen and consumpion, and expor rises even more. Consisen wih he rapid decline in employmen deviaions, he deviaions of oher macroeconomic indicaors also decline rapidly. By 2022, he deviaions are only slighly higher han hose in he unargeed policy. 5.3 Resuls for Perh under he RoWA- argeed policy Ineresingly, simulaion resuls show ha in he RoWA-argeed policy benefis no only he RoWA region, bu also Perh, and hence he Wesern Ausralia sae as a whole. This is because, as discussed earlier, Perh is he mos likely desinaion for people who leave RoWA. Table 2. GDP income and expendiure componens for RoWA under alernaive policies (Percenage deviaions from baseline) A. RoWA argeed policy 1.Real GDP Employmen (wagebill weighed) Capial sock Real privae consumpion Real invesmen Real public consumpion Expor volume Impor volume B. Unargeed policy 9.Real GDP Employmen (wagebill weighed) Capial sock Real privae consumpion Real invesmen Real public consumpion Real expor Real impor On he oher hand, RoWA is he mos likely desinaion for people who leave Perh. Census 2006 daa show ha beween he Census nigh and one year earlier, ou of 5.8% of people who lef RoWA, 3.98 percenage poin, or 68.4%, wen o Perh; and ou of 2.6% of people who lef Perh, 1.44 percenage poin, or 55.3%, wen o RoWA. Therefore, when he 20

21 wage rae fall sharply in RoWA under he argeed migraion program (Figure 6), more people will leave RoWA for Perh, and fewer people will leave Perh for RoWA. This largely explains he posiive conribuions of boh inerregional immigraion and inerregional emigraion o employmen deviaion in Perh, as repored in Figure 7. The posiive deviaion in Perh s employmen leads o changes in economic indicaors similar o hose in RoWA, only a a much smaller scale. Figure 7. Conribuion of populaion flows o he deviaion of WAP in Perh under he RoWA argeed policy, unless oherwise saed (Number of persons, deviaion from baseline) New Inersae immigraion Inersae emigraion Overseas emigraion WAP WAP, unargeed policy 6 Resuls for oher regions Figure 8 repors macro resuls for oher non-capial ciy regions in argeed simulaions. The paern of changes in he deviaion for employmen, real GDP and capial sock in all regions are similar o ha in RoWA. The differences in he magniude of he deviaions are largely deermined by he differences in he magniude of he shocks o he regional WAP as repored in column 4, Table 1. 21

22 Figure 8. Real GDP, Employmen and Capial sock in non-capial ciy areas under argeed migraion programs (Percenage deviaions from baseline) Res of New Souh Wales Res of Vicoria Real GDP Employmen Capial sock Real GDP Employmen Capial sock Res of Queensland Res of Souh Ausralia Real GDP Employmen Capial sock Real GDP Employmen Capial sock Res of Tasmania Res of Norhern Terriories Real GDP Employmen Capial sock Real GDP Employmen Capial sock For all regions, a once-off argeed migraion program does no seem o bring susained increase in economic aciviies in he long run. Over imes, new migrans move o oher regions, mosly o capial ciies, or go overseas. These resuls are consisen wih 22

23 opinions from DIAC, such as migraion is an unlikely remedy for susained populaion decline in regional Ausralia (Cully, 2010, p.12), or no credible policy will sop 90 per cen of fuure immigrans iniially or evenually seling in he capial ciies. (DIAC s Red Book briefing for Julia Gillard, cied in Bourke, 2011). 7 Concluding remarks Over a number of years he Ausralian Governmen has used regionallydiscriminaing migraion policy as an insrumen o arge a reducion in he naion s perceived over-urbanisaion. 21 The policy is implemened by a program of easing he graning of visas o skilled migrans inending o sele ouside of Ausralia s major populaion cenres. The policy is employed boh for permanen skilled visa classes and for emporary skilled visas. Over recen years here has been an increasing use of he emporary skilled visa, T457, in his way. This poses he quesion as o wheher a emporary (4-year) visa can be a successful regional developmen insrumen and o wha exen i can leave a developmen legacy in he argeed non-meropolian regions. To explore his quesion we inroduce a labour supply heory wih exensive migraion deail ino a muliregional dynamic CGE model, TERM. The model feaures 15 regions, 7 of hem non-meropolian and a labour supply specificaion providing a deailed reamen of labour choice behaviour by workers idenified by skill class, employmen saus, indusry, region, age group and visa saus. The model is parameerised o capure: he underlying ransiions beween labour caegories (such as hose beween visa caegories); and worker preferences regarding choices beween occupaions and regions of employmen. Our simulaions show ha regional argeing does generae shor-run improvemens for he argeed regions in variables which are ofen used as regional developmen indicaors, such as employmen and real GDP. However, labour-marke displacemen effecs ensure ha, even in he year of he increased inake, he expansion in he populaion of he argeed region is significanly below he size of he visa program. This resuls from a depressing of he argeed region s real wage rae consequen upon he increased regional labour supply, and his in urn causes a reducion in ne inerregional migraion o he argeed region. Over ime wih an ongoing, hough diminishing, impac on ne inerregional migraion and a porion of 21 The problem of migrans o sele mainly in big ciies is by no means unique o Ausralia. See a review in Hugo and Monren-Alegre

24 migrans reurning overseas, regional developmen effecs quickly move owards hose emanaing from an unargeed inake. The policy does, however, resul in some increased invesmen in argeed regions for a few years, wih a small residual regional simulus effec sill presen afer a decade. 24

25 References ABS (2006), Census 2006: Geographical area, People characerisics (age, labour force saus, occupaion, highes non-school qualificaion, ehniciy, region of residence one year ago), Table Builder, accessed 26 Sepember 2011 a hp:// ABS (2008a) Labour Force saus and oher characerisics of recen migrans, Nov 2007, ca. No Rerieved 19 Ocober 2011 a <hp:// CD36/$File/62500_nov% pdf>. ABS (2008b) Labour Mobiliy, February 2008, Expanded CURF, ABSDL. Daa compilaion based on use of ABS CURF daa. ABS (2009). Ausralian Naional Accouns: Inpu-Oupu Tables Elecronic Publicaion, Final, ca. No , Canberra. Available a: hp:// ABS (2010a) Characerisics of recen migrans, Nov 2010, Expanded CURF, ABSDL. Daa compilaion based on use of ABS CURF daa. ABS (2010b) Labour Mobiliy, February 2010, Expanded CURF, ABSDL. Daa compilaion based on use of ABS CURF daa. ABS (2010c) Trends in hours worked, ca Ausralian Labour Marke Saisics. Accessed 4 Ocober 2012 a hp:// 3FA6CA A9D9?OpenDocumen ABS (2010d) Deahs, Summary, Sae and erriories 2000 o Excel spreadshee, ca. No , viewed 1 Ocober 2011, <hp:// men.> ABS (2011) Ausralian Demographic Saisics, Sep Tables 51 58: Esimaed Residen Populaion by single year of age, for 8 saes and erriories. Excel spreadshee, ca. No rerieved 23 April 2012 from <hp:// ndocumen>. ABS (2012a), LM1 - Labour Force Saus by Sex, Age, Marial Saus, Capial Ciy/Balance of Sae, daa cube: SuperTABLE, ca. no , rerieved 26 April 2012 from hp:// 2?OpenDocumen ABS (2012b), UM3_Apr01 - Unemployed Persons by Sex, Sae, Age, Duraion of Unemploymen Caegories, April 2001 onwards, daa cube: SuperTABLE, ca. no , rerieved 26 April 2012 from <hp:// 12?OpenDocumen>. ABS (2012c), NM1_Apr01 - No in he Labour Force Saus by Sex, Sae, Age, daa cube: SuperTABLE, ca. no , rerieved 26 April 2012 from 25

26 2?OpenDocumen. ABS (2012d) E08_Aug96_ANZSICO06: Employed persons by occupain uni groups, sex, sae, saus in employmen, daa cube: SuperTABLE, ca. no , rerieved 26 April 2012 from 12?OpenDocumen>. ABC news (2012) Anger over foreign worker impor plan, online aricle, accessed 10 Augus 2012 from hp:// Access Economics (2002a). The Impac of Sponsored Temporary Business Residens on Ausralia's Living Sandards. Canberra: Deparmen of Immigraion, Muliculural and Indigenous Affairs. Access Economics (2002b). The Impac of Sponsored Temporary Business Residens on he Commonwealh Budge. Canberra: Deparmen of Immigraion, Muliculural and Indigenous Affairs Armsrong, H. and Taylor, J. (2000), Regional Economics and Policy, 3 rd Ediion, Blackwell, Oxford. Birrell, B., Heard G. (2012) The immigraion program: record number, ciy-bound. CPUR (Cenre for Populaion and Urban Research) Bullein, May Bourke W. (2012). Creaiviy, no slogans, is he answer o populaion dilemma, The Drum Opinion, 16 May 2011, accessed 1 Ocober 2012 a hp:/ Bowen, C., (2012a). Skilled migraion benefis and reforms: Address o he Naional Press Club. Deparmen of Immigraion and Ciizenship, Miniser s press release, 25 May Accessed 13 Augus 2012 a hp:// Cameron, R. (2011) "Responding o Ausralia's Regional Skill Shorages Through Regional Skilled Migraion," Journal of Economic and Social Policy: Vol. 14: Iss. 3, Aricle 4. Crean, S. (2012). Sronger regions, sronger naion, Miniserial saemen from he Miniser for Regional Ausralia, Regional Developmen and Local Governmen, in he Budge 1212/13. Accessed 6 Aug 2012 from hp:// Cully, M. (2010). The conribuion of migrans o regional Ausralia, Issues Paper, Deparmen of Immigraion and Ciizenship, Canberra. DIAC Deparmen of Immigraion and Ciizenship. DIAC (2011) Socks of visa holders, Inernal projec daa, received from DIAC December DIAC (2012a). Temporary Business (Long Say) (subclass 457) Visa. Informaion bookle. Accessed 6 Augus 2012 from hp:// DIAC (2012c). Reforms o he Permanen Employer Sponsored Visa Program. Accessed 6 Augus 2012 from hp:// 26