A Comparative General Equilibrium Analysis of the Estonian Labour Market. Alari Paulus Andres Võrk Kari E. O. Alho (ETLA) January 2007

Transcription

1 28 A omparatve General Equlbrum Analyss of the Estonan abour Market Alar Paulus Anres Võrk Kar E. O. Alho (ETA January 2007 Ths paper analyses the mpact of several tax-beneft reforms on the employment of fferent skll groups uner varous wage formaton systems. We aapt a general equlbrum moel ntally evelope for the Fnnsh economy by Alho (2006, combnng t wth elements from Bovenberg et al. (2000 an Hnnosaar (2004a, b. Polcy smulatons emonstrate the mportance of wage settng for the outcome. In the case of Estona, market etermne wages outperform bargane wages, whch coul mply an mportant fference between the EU-5 countres an the new member states n general. Atonally, the labour supply of the low-sklle s most effectvely ncrease by lowerng the margnal ncome tax rate, whch combne wth strateges ncreasng the overall employment, coul potentally mprove the labour market poston of those wth lower sklls. Ths s a part of the proect Tax/beneft systems an growth potental of the EU (TAXBE, Proect no. 8-T , fnance by the European ommsson uner FP 6 of Research. Keywors: computable general equlbrum moels, labour tax polcy, wage structure JE classfcaton: D58, E62, J3

2 A OMPARATIVE GEERA EQUIIBRIUM AAYI OF THE ETOIA ABOUR MARKET Alar Paulus, PRAXI Anres Võrk, PRAXI Kar E. O. Alho, ETA Abstract: Ths paper analyses the mpact of several tax-beneft reforms on the employment of fferent skll groups uner varous wage formaton systems. We aapt a general equlbrum moel ntally evelope for the Fnnsh economy by Alho (2006, combnng t wth elements from Bovenberg et al. (2000 an Hnnosaar (2004a, b. Polcy smulatons emonstrate the mportance of wage settng for the outcome. In the case of Estona, market etermne wages outperform bargane wages, whch coul mply an mportant fference between the EU-5 countres an the new member states n general. Atonally, the labour supply of the low-sklle s most effectvely ncrease by lowerng the margnal ncome tax rate, whch combne wth strateges ncreasng the overall employment, coul potentally mprove the labour market poston of those wth lower sklls. Keywors: computable general equlbrum moels, labour tax polcy, wage structure JE classfcaton: D58, E62, J3 orresponng author: Anres Võrk PRAXI enter for Polcy tues Estona pst 5a, Tallnn 043, Estona e-mal: anres.vork@praxs.ee Ths s a part of the proect Tax/beneft systems an growth potental of the EU (TAXBE, Proect no. 8-T , fnance by the European ommsson uner FP 6 of Research. We thank Karsten taehr an all the semnar partcpants at PB for helpful comments. All remanng errors are ours. 2

3 ontents. Introucton The moel Househols Frms Wage formaton Government albraton Polcy smulatons Descrpton owerng the margnal ncome tax rate Increasng the ncome tax allowance owerng employers socal securty contrbutons Increasng the replacement rate onclusons... 9 References Appences... 2 Appenx. A comparson of GE moels... 2 Appenx 2. abour supply Appenx 3. abour cost Appenx 4. Wage barganng

4 . Introucton The European Unon countres are ncreasngly concerne about ther compettveness n the global market. One of the central ssues s relate to the functonng of the labour market an socal protecton systems. In frequent comparsons of the U an the EU labour market, the latter has been consere more regulate an rg, whch agan has been assocate wth hgher unemployment rates. On the other han, labour n Europe enoys hgher socal protecton stanars. Uner the pressure of global processes, current trens are towars austments n tax-beneft systems, whch coul ncrease work ncentves an mprove flexblty of labour market wthout scalng back socal protecton too much (arone an alomäk, 200. Also the re-launche sbon trategy an the unerpnnng ntegrate guelnes avocate more employment frenly tax-beneft systems. The enlargement of the EU n 2004 ntrouce new member states, whch, havng relatvely ecentralse labour markets, also contrast wth the EU-5 countres. There are some concerns that ths coul lea to socal umpng. In ths context, the new member states have a lemma as to whch way to procee contnung the marketorente flexble approach or shftng to a more centralse barganng an protectve system. There s some emprcal evence that a bell-shape relatonshp exsts between the centralsaton of wage barganng an the unemployment level (almfors an Drffll, 988, possbly makng choces n-between relatvely unfavourable. In ths paper we take Estona as one example of the new member states an try to answer whether t woul be benefcal to mplement a tax-beneft system more akn to those foun n the ol EU countres. Estona s a small open economy conuctng lberal economc an tax polcy. Recent an on-gong tax-beneft reforms ame at lowerng the ncome tax buren an to ncrease unemployment an subsstence benefts represent a goo opportunty to moel the outcome uner varous wage formaton hypotheses. We aopt a computable general equlbrum (GE moel ntally evelope for Fnnsh economy as a part of the TAXBE proect, see Alho (2006, but also wth elements from the moels of Bovenberg et al. (2000 for Dutch an Hnnosaar (2004a, b for Estonan economy. 4

5 The man features of the moel are the followng. There are two proucton factors captal an labour, the latter ve further nto three skll groups base on eucatonal attanment. Frms are symmetrc an prouce one homogenous goo. The goos market s characterse by monopolstc competton, mplyng postve profts for frms. The foregn sector s not explctly moelle, omestc frms compete wth foregn frms n the nternatonal market an t s assume that the omestc prce level of goos equals the nternatonal prce level. Househols earn labour ncome, receve strbute profts an unemployment benefts. Ther utlty epens on lesure, prvate consumpton, on whch all the ncome s spent, an publc consumpton. Government has a passve role of spenng all tax ncome on unemployment benefts an publc consumpton. Tax revenue s generate by ncome taxes on labour an captal an employers socal securty contrbutons. Throughout the moel, functons wth constant elastcty of substtuton (E have been use (e.g. for proucton, aggregaton of proucton factors, utlty. Three fferent structures of wage formaton are moelle. Frst, fxe wages, whch n case of a tax-beneft polcy change woul reflect the frst reacton n the (very short run. econ, market etermne wages, whch may correspon to the Estonan case uner current crcumstances n the meum run. (We o not conser the long run as captal s hel fxe. Thr, wage barganng by each skll group, representng a more EU-orente hypothetcal case. Overall, labour supply an wage barganng are moelle n the manner of Bovenberg et al. (2000 an Hnnosaar (2004a, b, whle the proucton se an other wage formaton schemes (fxe an market etermne are moelle as n Alho (2006. A more etale comparson of the prevous moels an the current approach s presente n Appenx. General equlbrum effects of Estonan tax-beneft system have not been extensvely researche. To our knowlege, there are no prevous stues apart from Hnnosaar (2004a, b. ompare to the latter, we conser several alternatve wage formaton systems. We also ntrouce captal as a proucton factor, although fxe, an employ more recent ata. Atonally, havng avalable a smlar moel to the Fnnsh case allows to compare tax-beneft effects on employment n an ol an a new member state, 5

6 where the coverage of wage barganng ffers notably; 90% an 20-30%, respectvely, n 2003 (European Inustral Relatons Observatory, Moellng several skll groups allows us also to analyse separately the stuaton of low-sklle labour, whose employment rate s partcularly low (see Fgure. Fgure. Employment rate, unemployment rate an eucatonal level n Estona, Employment rate, % 25 Unemployment rate, % below upper seconary ource: tatstcal Offce of Estona upper seconary tertary The rest of the paper s structure n the followng way. ecton 2 scusses the moel. ecton 3 presents the ata an the results of the calbraton. ecton 4 consers fferent polcy smulatons (lowerng the margnal ncome tax rate, ncreasng the ncome tax allowance, lowerng employers socal securty contrbutons an ncreasng the unemployment beneft replacement rate. ecton 5 conclues. 6

7 2. The moel 2.. Househols The workforce conssts of three types of househols fferng n sklls base on eucatonal attanment: low-sklle, sklle an hgh-sklle (,2, 3. There are M househols n each skll category who maxmse utlty subect to a buget constrant an a tme constrant. The utlty of househol wth skll level s a functon of prvate consumpton, lesure V an publc consumpton G. The latter enters n an atvely separable way, therefore havng no effect on labour supply ecsons 2 : H h(, V g( G ( + The utlty sub-functon s the E type: (, ( h + ( ( V V (2 where s the elastcty of substtuton between consumpton an lesure an s a strbuton parameter mplyng the relatve weghts whch househols place on consumpton an labour supply. A househol s buget constrant s ( t W P (3 where t s the average tax rate on labour ncome (a functon of labour ncome, tax rates an tax free allowance, W s the gross wage rate, s labour supply an P s the prce level. Househols also receve captal ncome from strbute profts, however t s assume that they are not able to antcpate t ex ante an therefore t oes not enter the buget constrant here. The tme enowment s normalse to unty, V,.e. the unt of tme enowment (an labour supply coul be consere as a full-tme ob. Ths secton follows Bovenberg et al. (2000 an Hnnosaar (2004a, b. However, there s an mportant fference n the aggregate labour supply equaton. ee Appenx 2 for the etals. 2 The utlty from publc consumpton for a skll group s erve as g( G g G, where g s the share of the skll group n publc consumpton. The latter s an exogenous varable. 7

8 The total labour supply of skll group s M tmes the labour supply househol: of one A M mt P + ( m W P ( m W P m W P ( (4 where m s margnal tax rate on labour ncome an T A s the tax allowance. Househols ece the optmal amount of labour supply, however, the partcpaton ecsons are exogenous. Total labour supply of skll group mples that n the en househols from that skll group are suppose to supply all ther labour n the moel an the rest are consere as nactve. In ths framework frms ece employment E, whch therefore equals the labour emane. ome househols are unemploye an receve nstea of labour earnngs unemployment benefts B (also taxe. Aggregate (realse prvate consumpton s equal to labour ncome, unemployment benefts an captal ncome, all net of taxes: B [( t W E + ( t BU ] + ( m P Π (5 where U s the number of unemploye (also corresponng to full-tme obs an Π s aggregate captal ncome. ote that the margnal tax rate s constant an not epenent on ncome type or skll group. In fact, all ncome s taxe n a common framework an tax allowance s n prncple apple to total ncome. However, t s assume that the tax allowance s assgne to ether labour ncome or unemployment benefts an both excee that, so that any atonal ncome,.e. captal ncome, must be taxe n full extent. The fferent average tax rates for labour ncome an unemployment benefts are ue to the fferent unerlyng amounts Frms Ths secton s base on Alho (2006. There s monopolstc competton n the economy between entcal frms. Frms use two aggregate proucton nputs captal an labour to prouce one homogenous goo. The aggregate proucton functon s a E functon mplyng constant returns to scale: 8

9 σ σ [ αk + ( α ] σ Q A (6 where A s total factor prouctvty, K captal stock, aggregate effectve labour nput, α a strbuton parameter an ( σ the substtuton elastcty between captal an aggregate effectve labour. The aggregate eman Q * n the nternatonal market epens on the nternatonal prce level P, Q P D ε b(, where ε > 0. Frms maxmse profts, whch assumng ournot competton leas to the eman for effectve labour nput: P Q Q P + Q ( Q P Q (7 where Q s the margnal prouctvty of aggregate labour, Q Q h the market share of omestc frms an ( s the aggregate unt labour cost. Ths can be rewrtten as: P ( h Q ( where ( ε ε (8 h s the nverse of the mark-up. Ths etermnes the employment. The captal stock remans fxe as we concentrate only on the short-meum run. Aggregate effectve labour nput s a E aggregate of labour emane n each skll group : [ ( ] e, (9 Here, e s the relatve effcency/prouctvty parameter of the skll group wth e fxe to unty an ( s the substtuton elastcty of skll groups. The gross wage rate s W, but the corresponng cost to the employer,.e. the proucer wage, s ( ( + v, where v s the rate of employers compulsory socal securty W contrbuton. ettng up an solvng a cost mnmsaton problem gves us the relatve eman for the fferent skll groups (see Appenx 3 for the ervaton: ( ( e e an the aggregate unt labour cost:, 2, 3 (0 9

10 ( ( ( e 2.3. Wage formaton mlarly to Alho (2006 we conser several wage formaton hypotheses. Frst, fxe wages reflectng the frst reactons n the (very short run after a tax-beneft polcy change. econ, market etermne wages corresponng to a flexble labour market wth market clearng n the meum run (captal stll fxe. Thr, we conser a case of wage barganng. The frst two settngs are perhaps closer to the actual stuaton n the Estonan labour market wthn the respectve tme horzon; the last one s somewhat more hypothetcal, reflectng the ol EU member states. The case of fxe wages s the easest to moel, requrng only the relevant constrant n the moel. Market etermne wages nee an atonal restrcton relate to unemployment, whch otherwse woul not exst n the moel. Here, unemployment rates have been fxe uner market etermne wages. Bargane wages are etermne by a rght-to-manage moel, whch s as follows, beng a combnaton of approaches by Hnnosaar (2004a, b an Alho (2006. The employers organsaton an three unons each representng workers of one skll group bargan over wages an employers etermne employment. Ths mples maxmsng the followng ash functon: Ω Λ Γ β β (2 where ( + v W E ( K Λ PQ( K, ρ + (3 an B [( t W ( t B ] 0. 5 Γ E (4 0.5 Λ stans for the total proft of the employers organsaton an Γ represents the utlty of the unon; ρ enotes the rate of return on captal an the eprecaton rate. It s assume that the unon attaches equal weght to employment an the surplus from 0

11 workng,.e. the real consumer wage less the reservaton wage what s smply the aftertax unemployment beneft. The parameter β enotes the relatve barganng power of the employers. olvng the ash functon gves the followng wage equaton of skll group (see Appenx 4 for the ervaton: W β B 0.5 ( + β + [ W E ( ρ + K ] ( β P Q( K, ( v + ( + v E ( + β ote that bargane wages o not epen on the margnal tax rate. Ths s ue to the fact that wages an benefts are taxe n the same manner, applyng the same margnal tax rate an tax allowance. Ths mples that f the employers organsaton ctates the outcome, β, the wage s suppresse own to the reservaton wage,.e. the (5 unemployment beneft, W B. In case a labour unon omnates the barganng, β 0, all the proucton surplus s attrbute to wage ncome: W ( + v ( + v E P Q( K, W E ( ρ + K ( Government The government runs a balance buget, spenng collecte tax revenues on unemployment benefts an publc consumpton. There s a unversal ncome tax apple on all ncome, nclung unemployment benefts, wth constant margnal tax rate. Yet, ncome taxaton s progressve ue to ncome tax allowance. It s apple to ether labour earnngs or benefts, epenng on whch of these a househol s recevng, whch are assume to be suffcently large so that the tax allowance coul be utlse fully. Therefore, the rest of the ncome, namely captal ncome n the form of strbute corporate profts, s subect to full taxaton. The average tax rate on labour earnngs coul be enote as t A ( T m W (7 W an the average tax rate on unemployment benefts as

12 t B A ( T m B (8 B where T A s the amount of tax allowance. Atonally, there s a compulsory socal securty contrbuton on employers wth a constant rate of v. o value-ae tax s use. Government uses ts resources to fnance unemployment benefts an the rest s spent on publc consumpton. Unemployment benefts are nexe to the gross wage rate, B rw, where r s the pre-tax replacement rate. Overall, the government buget s: B [ v + t W E ( t BU ] + mπ PG ( (9 3. albraton The moel has been calbrate to the Estonan economy n Input ata an parameters are shown n Table. abour market ata an gross wage rates have been erve from the abour Force urvey (tatstcal Offce of Estona, There s a populaton of 830. thousan of persons n workng age (6 to penson age. Ths s ve nto three skll groups base on eucatonal level (below upper seconary, upper seconary an tertary: low-sklle, sklle an hgh-sklle. abour force partcpaton rates across the skll groups are, respectvely, 42.4%, 76.6% an 88.2%, an unemployment rates 20.8%,.0% an 5.4%. The gross wage rate for sklle an hgh-sklle s hgher by respectvely 9% an 33% than that of low-sklle workers. Although hourly ata woul have been preferre, t was not avalable, therefore labour market ata s presente n persons an wage rates are unauste for part-tme workng. Total proucton s 4.5 bllon of EEK (tatstcal Offce of Estona, 2006 an the level of captal stock s 2 bllon of EEK, both n nomnal values of The captal stock s estmate usng the Perpetual Inventory Metho, where the captal stock equals ntal captal stock an past nvestments less the eprecaton. It s assume that the ntal captal stock was.5 tmes the resual value of fxe assets on enterprse balance sheets at the en of 992. The captal stock was then calculate assumng annual eprecaton rate of 0% an usng fxe nvestment eflator. The same eprecaton rate was also use later for the calculaton of aggregate profts. The moel s set up for a 2

13 small open economy, where omestc prce level follows the nternatonal prce level. The latter has been fxe to unty. The parameters charactersng the tax-beneft system n the moel are the followng: the margnal ncome tax rate m s 26%, the annual ncome tax allowance T A s 6,800 EEK an the socal securty contrbuton rate for employers v s 33%. Uner current unemployment nsurance, unemployment assstance an socal assstance system, the gross replacement rate s 50% for the frst 00 ays, 40% for the followng 80 ays an after that about 5% for effectvely unlmte uraton. Here t s assume that all unemploye are enttle to the hghest replacement rate. Therefore, unemployment benefts gross replacement rate r n the moel s 50%. Table. Input ata for moel calbraton abour market a Populaton M 48.7 M M Employment E 50.0 E E Unemployment U 3. U U Gross wage rates b W 53.0 W W atonal accounts c Q 4.5 K 2.0 Tax-beneft rules M 0.26 T A 6.8 b ν 0.33 r 0.5 Parameters α σ a Thousans of persons (n workng age 6 to penson age. b Thousans EEK (annually. c Bllons EEK The substtuton elastcty between the skll categores, 2, an the substtuton elastcty of captal an aggregate labour, σ 0. 8, are erve from Alho (2006. Although the moels an the unerlyng economes are fferent, we use those as no estmates for Estonan economy are currently avalable. The strbuton parameter α n the proucton functon s set to 0.4. The substtuton elastcty of consumpton an lesure s set equal to 2, followng Hnnosaar (2004a, b. albrate ata mples that n the ntal equlbrum, total factor prouctvty s 4.06 an the rate of return on captal s 2.4%. The mark-up factor s.92, whch s relatvely hgh, on the other han there s no value-ae tax present n the moel. albrate parameters an some mportant varables for the skll groups are presente n Table 2. 3

14 Table 2. albrate parameters an varables by the skll types klle owsklle Hghsklle Parameters The relatve effcency of a skll group, e The barganng power of employers, β The strbuton parameter n the utlty functon, The share of a skll group n publc consumpton, g Varables The share of a skll group n captal ncome Average ncome tax rate on labour earnngs, t 7.8% 9.% 20.8% B Average ncome tax rate on unemployment benefts, t 9.5% 2.% 5.6% There s only a margnal fference between the relatve effcency of sklle an hghsklle labour. The group of sklle has even stronger barganng power compare to the hgh-sklle. Each skll group s share n publc consumpton s assume to be proportonal to the number of respectve househols an therefore beng exogenous. The share n captal ncome s proportonal to the aggregate wage earnngs. Fnally, the values of the strbuton parameter n the utlty functon reflect that those wth hgher sklls attrbute a larger weght to consumpton. Ths s ue to hgher unemployment among lower skll groups, whch n ths framework translates (partly nto stronger preferences for lesure. 4. Polcy smulatons 4.. Descrpton There are four polcy scenaros evaluate uner all three wage systems, altogether up to 9 fferent smulatons (some prove the same results. The followng polcy changes are consere: lowerng the margnal ncome tax rate, 2 ncreasng the ncome tax allowance, 3 lowerng employers socal securty contrbutons, 4 ncreasng the replacement rate. 4

15 All polcy smulatons are fnance by an ex-ante reucton n the level of publc consumpton by 0.5%. In terms of tax-beneft parameters ths mples that (a margnal ncome tax rate s lowere from 0.26 to 0.222, (b tax allowance s ncrease from 6,800 to 7,577 EEK per year, (c employers socal tax rate s ecrease from 0.33 to 0.327, an ( unemployment beneft replacement rate s ncrease from 0.5 to The two frst polcy shfts mtate actual ncome tax reforms n 2003 an 2005, whch wll reuce the margnal ncome tax rate from 26% to 20% an ouble the annual tax allowance from 2,000 to 24,000 EEK once fully mplemente. However, here we o not follow the actual polcy changes n ther exact magntue as the fferent smulatons are balance n fscal terms n orer to retan comparablty. The latter two scenaros are more hypothetcal, carre out as the man alternatves. o smulaton of targete polcy changes n the form of e.g. a hgher replacement rate for the low-sklle have been unertaken. The results are presente n Table 3. 5

16 Table 3. mulaton results, percentage changes Polcy scenaros ( ower margnal ncome tax rate (2 Increase ncome tax allowance (3 ower employers socal securty contrbuton (4 Increase replacement rate Target varables F, B M F, B M F M B F, M B Proucton Prvate consumpton Publc consumpton ocal welfare low-sklle sklle hgh-sklle abour supply low-sklle sklle hgh-sklle Employment low-sklle sklle hgh-sklle Unemployment low-sklle sklle hgh-sklle Gross wage rate - low-sklle sklle fxe -0.6 fxe 0.0 fxe fxe.6 - hgh-sklle Unemployment rate a low-sklle fxe fxe fxe - sklle hgh-sklle ote: F fxe wages, M market etermne wages, B wage barganng. a Absolute changes n percentage ponts owerng the margnal ncome tax rate We can nee observe ncrease prvate consumpton wth a lower margnal ncome tax rate. On the other han, tax revenues eclne an therefore publc consumpton s smaller. The overall mpact on welfare epens on the socal welfare functon use. In ths case prvate an publc consumpton share the same strbuton parameter an 6

17 socal welfare also ncreases. Across the skll groups, more sklle persons gan more as they have hgher preference for ther consumpton (see the comments n ecton 3. In case of fxe (gross wages only consumer wages ecrease, proucer wages o not react an therefore proucton an labour eman wll not be affecte. Hgher consumer wages ncrease the labour supply. But employment s etermne by the frms an constant (as the proucer wages are fxe, an the ncrease labour supply translates one-to-one nto hgher unemployment. Market etermne wages are more flexble an the gross wage rate must ecrease for the unemployment rates to be unchange. Therefore, the gans from reuce labour taxes are share between employers an employees. As proucer wages ecrease, more labour s emane an hgher employment levels are attane. In the new equlbrum, unemployment levels are only slghtly hgher an unemployment ratos are the same. Yet, proucton also ncreases an therefore prvate consumpton an socal welfare ncrease more than uner fxe wages. The wage barganng case has the same outcome as fxe wages because the bargane wages o not epen on the margnal tax rate. The lower margnal tax rate only ncreases the labour supply, whch has no (rect effect on employment Increasng the ncome tax allowance A larger ncome tax allowance also ncreases househols sposable ncome. However, the effect on the labour supply s exactly the opposte now. Uner fxe wages ths transforms nto lower unemployment as proucton an employment are not affecte. Prvate consumpton ncreases a lttle, but the overall effect on socal welfare s only margnal. In case of market etermne wages the ecrease n labour supply s smaller. Ths n turn requres a eclne n employment an a lower gross wage rate. Proucton also ecreases an the overall effect on socal welfare remans negatve. Wage barganng yels agan the same results as fxe wages owerng employers socal securty contrbutons Reucng the labour tax buren by lowerng employers socal securty contrbutons has a rect effect on labour eman an employment, but not on labour supply. ower 7

18 labour costs ncrease proucton an employment. Agan, as there s no mechansm lnkng labour eman an supply, the latter remans unchange wth fxe wages an unemployment eclnng. Hgher employment an proucton ncrease prvate an publc consumpton, therefore rsng socal welfare as well. Fxe unemployment rates are attane va hgher consumer wages an ncrease labour supply, whch makes ths polcy scenaro less favourable uner market etermne wages. Proucton ncreases only margnally, prvate consumpton ncreases less than uner fxe wages an publc consumpton even ecreases. Overall, socal welfare s somewhat hgher. Ths tme wage barganng leas to a fferent result compare to fxe wages, n fact beng closer to the outcome of market etermne wages. The ntal ecrease n proucer wages transforms to hgher gross wage rates. In the new equlbrum, proucer wages, employment an proucton have returne to ntal levels. Hgher consumer wages attract atonal labour supply causng unemployment levels to rse. Overall, there s a small ncrease n prvate consumpton, a small ecrease n publc consumpton an only a margnal mprovement n socal welfare Increasng the replacement rate The ncrease replacement rate combne wth ether fxe wages or market etermne wages only mples slghtly hgher prvate consumpton an socal welfare. The reason s that nether frms nor househols take nto account the extent of unemployment benefts n ther optmsng behavour. It oes not affect labour costs for the frms an househols supply labour wthout conserng the outse opton. Unemploye househols are exogenously ece an therefore they o not take nto account unemployment benefts when choosng optmal labour to supply. Uner wage barganng, hgher unemployment benefts rectly ncrease gross wage rates, whch n turn ncreases labour supply. Hgher labour costs lower frms eman for labour an therefore the employment level. As replacement rate s unform across the skll groups, the percentage changes n the gross wage rates an employment are also the same. All ths results n sgnfcantly hgher unemployment, ecrease proucton, lower prvate an publc consumpton an lower socal welfare. 8

19 5. onclusons The polcy smulatons consere show that alternatve ways to stmulate the labour market can lea to very fferent outcomes, e.g. on labour supply an unemployment. An mprovement n terms of househols sposable ncome mght even turn out to be welfare reucng n the new equlbrum. The effects of polcy changes also vary uner fferent wage formaton schemes lowerng the margnal ncome tax rate s for example most effectve n enhancng prvate consumpton an socal welfare uner market etermne wages whle a reucton n the socal tax rate works most successfully uner fxe wages. A combnaton of lowerng margnal ncome tax rate an ncreasng tax allowance, bascally the recent tax reform, has a potental to ncrease proucton an socal welfare wthout ncreasng unemployment rates uner market etermne wages. Overall, the results are often smlar to Hnnosaar (2004b an Alho (2006 on whch our moel s base. In comparson wth the latter, the man fferences occur where our smulatons yel entcal results uner varous wage systems, hghlghtng some lmts of our moel. However, assumng that fferent wage formatons are relevant for Estonan an Fnnsh economes (market etermne wages an bargane wages, respectvely, we can stress the nee for fferent labour market an tax-beneft polces n fferent EU member states. omparng the polcy scenaros for Estona uner market etermne wages an wage barganng mples that market etermne wages outperform bargane wages, the latter representng more EU-lke wage formaton. Although no polcy scenaros targete at specfc skll groups were carre out, some mplcatons coul be stll note. The labour supply of low-sklle s most effectvely ncrease by lowerng the margnal ncome tax rate, val uner every wage scheme. ombnng ths n turn wth strateges mprovng employment n general, e.g. lowerng employers socal securty contrbutons, coul potentally mprove the labour market poston of those wth lower sklls. 9

20 References Alho, K. (2006. abour market nsttutons an the effectveness of tax an beneft polces n enhancng employment: a general equlbrum analyss. ETA Dscusson paper, o 008. Bovenberg,.; Graaflan, J.; Moo, R. e (2000. Tax reform an the Dutch labor market: an apple general equlbrum approach, Journal of Publc Economcs, o 78, pp almfors, ; Drffll, J. (988. Barganng tructure an Macroeconomc Performance, Economc Polcy, 6, Aprl, pp arone, G.; alomäk, A. (200. Reforms n Tax-Beneft ystems n Orer to Increase Employment Incentves n the EU, EFI, Economc Paper, o. 60. European Inustral Relatons Observatory, hanges n natonal collectve barganng systems snce 990. [ ast accesse: ov, Hnnosaar, M. (2004a. Estonan abor Market Insttutons wthn a General Equlbrum Framework, Eest Pank, Workng Paper, o 5. Hnnosaar, M. (2004b. The Impact of Beneft an Tax Reforms on Estonan abor Market n a General Equlbrum Framework, Unversty of Tartu, Workng Paper, o 3. abour Force urvey atabase. tatstcal Offce of Estona, tatstcal Offce of Estona, tatstcal Database. [ ast accesse: Feb 6,

21 Appenx. A comparson of GE moels Moel Bovenberg et al. (2000 Tax reform an the Dutch labor market: an apple general equlbrum approach, Journal of Publc Economcs, o 78, 2000, pp Backgroun Househols A smplfcaton of MIMI (a larger apple general equlbrum moel for the Dutch economy. Three types of househols: captalsts, unsklle an sklle househols. aptalsts o not supply labour an receve all proft ncome. Partcpaton ecsons are exogenous, househols choose the number of workng hours. Househols of each skll type maxmse utlty, subect to a buget an a tme constrant (publc consumpton atvely separable. Hnnosaar, M. (2004 a Estonan abor Market Insttutons wthn a General Equlbrum Framework, Eest Pank, o 5, b The Impact of Beneft an Tax Reforms on Estonan abor Market n a General Equlbrum Framework, Unversty of Tartu, Workng Paper, o 3, Base largely on Bovenberg et al. (2000, wthout nformal labour market, ob matchng an hrng costs. The wage formaton for hgh-sklle workers s base on the effcency wage concept. As n Bovenberg et al. (2000, except no hrng costs an tax progressvty ue to tax allowance, not tax cret. Alho, K. (2006 abour market nsttutons an the effectveness of tax an beneft polces n enhancng employment: a general equlbrum analyss. ETA Dscusson paper, o 008. ome smlar elements wth Bovenberg et al. (2000 (ob matchng an hrng costs. Atonal nput n proucton captal. Three types of skll groups (lowsklle, sklle an hgh-sklle, each ve further as employe on market-base terms, employe on subsse terms, unemploye an those outse the labour force. Invuals maxmse utlty subect to a buget an a tme constrant: The current paper A omparatve General Equlbrum Analyss of the Estonan abour Market. PRAXI, Workng Paper. Base on Hnnosaar (2004a, b an Alho (2006. As n Hnnosaar (2004a, b. Three skll groups nstea of two. Workers receve strbute corporate profts, although t s not consere n labour supply (not antcpate. 2

22 H u(, V + h( G, s. t. ( TA ( D W P c V whch yels the labour supply (n terms of hours. H m u( m u( m, Z θ θ θ θ θ m θ θ m θ ( + ( ( Z m Z m ( TA W, Z m m m + g( G, P c m s. t. H H (,, G, * P F U ( t W + ( t bu + O + Tr s. t. Decsons to partcpate n the labour market an to supply labour are consere smultaneously. Unemployment benefts are means-teste. B Frms Two types of omestc frms, accorng to the type of labour employe (unsklle, sklle. The number of frms s fxe. A lnear proucton functon, Y h. Proft maxmsaton mples uner monopolstc competton output prces (mark-up over margnal costs: W u, s P, ε h... ommotes prouce by entcal frms are aggregate nto two composte commotes an those are aggregate further nto overall proucton. The optmal allocaton of two composte commotes s an mplct eman functon for sklle an unsklle labour. As n Bovenberg et al. (2000, except proucer wage nclues employer s compulsory socal securty contrbutons. P W ( + Ts ε h Ientcal monopolstc frms. E aggregate proucton functon s σ σ / [ αk + ( α ] σ Q A. Proft maxmsaton uner ournot competton mples aggregate labour eman: * ( hε Q ( tq P. In the long run captal stock also * austs: Q ρk. P Q abour nput n each skll group: [ ] χ χ / χ + a M. The relatve eman for two components (trough cost mnmsaton: χ χ M M ( + v ( s a. ( + u Aggregate labour nput s I [ ( e ] / an the As n Alho (2006, except no subsse labour. 22

23 Goos market abour market Government There are two aggregate goos on the hghest level: (a a E aggregate of omestc composte commotes prouce by sklle an unsklle workers, (b a composte of mporte commotes. Realse employment equals to labour emane an unemployment s the fference between labour supply an labour eman (both n terms of hours. Wage barganng for both skll type labour (a rght-to manage moel. The reservaton wage s a weghte average of opportunty wages n formal an nformal sector, ntroucng some real wage resstance,.e. hgher tax rates or hrng costs only partly borne by workers. Atonally, ob matchng wth hrng costs. Government runs a balance buget wth revenue from taxng labour ncomes an expenture on unemployment benefts an publc consumpton. The latter featurng the same composton As n Bovenberg et al. (2000. Determnaton of employment an unemployment as n Bovenberg et al. (2000. Wage barganng for low-sklle workers as n Bovenberg et al. (2000, except no nformal labour market, ob matchng or hrng costs. The reservaton wage s smply unemployment beneft. Hgh-sklle workers receve an effcency wage epenng (among others on unemployment beneft an unemployment. As n Bovenberg et al. (2000, except government revenues nclue also employers compulsory socal securty contrbutons. Tax progressvty ue to tax allowance not tax cret. relatve eman for each skll group s ( ( e e ( ( A sngle traable goo, both n prvate an n publc consumpton. Job matchng smlar to Bovenberg et al. (2000, establshng a lnk between labour eman an labour supply. Frms ece employment. everal wage formatons consere: a fxe wages, b market etermne wages, c wage barganng n each skll group, naton-we wage barganng. Balance buget. Tax revenues from ncome tax (wth progressve rates on labour ncome, unemployment benefts an captal ncome, socal securty tax an value-ae tax. Government As n Alho (2006. Three wage structures consere: a fxe wages, b market etermne wages, c wage barganng as n Hnnosaar (2004a, b. Balance buget. Government revenues from labour an captal taxaton (ncome an socal tax. o value-ae tax. Government expentures on unemployment benefts an publc consumpton. 23

24 Foregn sector Data an the key parameters mulatons an prce nex as prvate consumpton. The allocaton of foregn consumpton over omestc an foregn goos epens on the terms of trae. The market of omestc goos s n equlbrum. Dutch economy n 208. ubsttuton elastcty between skll groups.5, substtuton elastcty of consumpton an lesure 4, uncompensate wage elastcty 0.5, ncome elastcty , export elastcty -2. owerng margnal tax rate (benefts nexe to consumer wages 2 Hgher tax cret for all househols (benefts nexe to consumer wages 3 Hgher tax cret for all househols (benefts nexe to proucer wages 4 Hgher tax cret for unsklle workers (benefts nexe to proucer wages 5 Hgher tax cret for all workers, hgher margnal tax rate, sklle workers break Unemployment benefts combne subsstence beneft an unversal unemployment beneft. As n Bovenberg et al. (2000. Estonan economy n 200, mport structure from 997. ubsttuton elastcty between skll groups 0.5, substtuton elastcty of consumpton an lesure 2, Armngton elastcty an transformaton elastcty 2. Hnnosaar 2004a: Hgher unon barganng power,.e. hgher wage for low-sklle workers 2 Hgher unemployment benefts replacement rate for all workers 3 Hgher unemployment benefts replacement rate for hghsklle workers 4 Increasng tax allowance for all workers 5 Increasng tax allowance for low-sklle workers expentures on unemployment benefts, wage subses an publc consumpton. Deman for omestc proucts epens on the nternatonal prce level (gven for a small open economy. Otherwse not explctly moelle. Fnnsh economy n ubsttuton elastcty between skll groups 2, substtuton elastcty of captal an aggregate labour 0.8, strbuton parameter n proucton functon 0.4, substtuton elastcty of consumpton an lesure 0.5. owerng ncome tax rate a average (an margnal b margnal (only 2 owerng employers socal securty contrbutons for a all skll groups b low-sklle workers 3 Hgher wage subsy rates 4 ower unemployment beneft replacement rates 5 Reucng unemployment n a fully flexble labour market Fnance by ex ante ecrease n publc consumpton by 0.5% n Unemployment beneft s a fxe proporton of gross wage. As n Alho (2006. Estonan economy n ubsttuton elastcty between skll groups 2, substtuton elastcty of captal an aggregate labour 0.8, strbuton parameter n proucton functon 0.5, substtuton elastcty of consumpton an lesure 2. owerng margnal ncome tax rate 2 Increasng tax allowance 3 owerng employers socal securty contrbutons 4 Increasng replacement rate. Fnance ex ante ecrease n publc consumpton by 0.5%. 24

25 Results even ex ante (benefts nexe to proucer wages Fnance by ex ante reucton n publc consumpton (0.5% GDP. Paper emonstrates varous traeoffs facng tax reforms. mulatons hghlght n-work benefts as an effectve nstrument aganst unemployment. Hnnosaar (2004b, reforms fnance by ex ante ecrease n publc consumpton by 0.5%: owerng margnal tax rate 2-5 As n Hnnosaar (2004a. Reucng tax buren va targete ncrease n tax allowance has the most favourable mpact on the labour market. an 2, 0.07% n 3, 0.8% n 4. mulatons show that wage formaton s an mportant factor for employment enhancng polces. In some cases the expansonary effects are even the largest uner wage barganng. mulatons show that market etermne wages outperform bargane wages. The labour market poston of the low-sklle coul be mprove most effectvely by lowerng the margnal tax rate an employers socal securty contrbutons. 25

26 26 Appenx 2. abour supply Invuals maxmse the E sub-utlty functon ( ( (, ( + V V h (20 subect to the tme constrant V an the buget constrant P W t ( (2 ubsttutng constrants nto the utlty functon: ( ( ( ( max + P W t h (22 FO, note that A mt m W W t + ( ( : ( ( ( ( ( ( ( ( ( 0 ( ( ( A A m W P P mt m W P mt m W (23 olvng for labour supply 3 : ( ( ( ( + A m W P P mt m W ( (24 ( ( ( ( + ( m W P P m W m W P P mt A (25 Total labour supply s ( ( ( µ µ µ + P m W P m W mt M M A, (26 3 ote that ths s fferent compare to Bovenberg et al. (2000 an Hnnosaar (2004a, b. There a househol s labour supply was also expresse n terms of average tax rate, whch n return epens on househol s labour supply.

27 27 Appenx 3. abour cost The aggregate labour cost mnmsaton problem: ( mn (27 subect to ( [ ] e (28 The agrangan functon s: ( ( [ ] e λ ˆ mn (29 FO: ( ( [ ] ( 0 ˆ ( e e e λ (30 Dve by, ˆ : ( ( e e (3 um over ( n... an a one to both se for : ( ( ( ( e e (32 ubsttute (27 an (28 nto (32, solve for : ( e (33 Multply by ( an sum over n... ( ( e (34 olve for aggregate labour cost, : ( e (35

28 Aggregate unt labour cost ( : ( ( (36 e Appenx 4. Wage barganng The ash functon to be maxmse s Ω Λ Γ β β where Λ P Q( K, + v W E ( ρ + (38 an FO: ( K B [( t W ( t B ] (37 Γ E (39 Ω W + β Λ β [ ( + v E ] Γ β β β B ( β Λ Γ 0.5E [( t W ( t B ] ( m 0 A A ote that ( ( ( (40 t m T W t W m W + mt. The unons o not take nto account that benefts epen on wage (otherwse employees woul clam very hgh wage an be unemploye. mplfyng yels: B [( t W ( t B ] 0.5( ( Λ β ( + v E β m (4 β ( + v W E β ( + v B E βb W ( + β ( β ( + v W E ( β P Q( K, ( + v [ + ] W E K ( ρ [ P Q( K, + v W E ( ρ + K ] ( β ( ( + v E ( + β (42 (43 In case β then W B, f β 0 then ( + v ( + v E P Q( K, WE ( ρ + K W (44 28