NBER WORKING PAPER SERIES INDIRECT TAX INITIATIVES AND GLOBAL REBALANCING. Chunding Li John Whalley

Transcription

1 NBER WORKING PAPER SERIES INDIREC AX INIIAIVES AND GLOBAL REBALANCING Chundng L John Whalley Workng Paper NAIONAL BUREAU OF ECONOMIC RESEARCH 15 Massachusetts Avenue Cambrdge, MA 2138 March 212 We are grateful to the Ontaro Research Fund and to the Centre For Internatonal Governance Innovaton Waterloo for fnancal support, and to a semnar group at UWO for comments. he vews expressed heren are those of the authors and do not necessarly reflect the vews of the Natonal Bureau of Economc Research. NBER workng papers are crculated for dscusson and comment purposes. hey have not been peerrevewed or been subject to the revew by the NBER Board of Drectors that accompanes offcal NBER publcatons. 212 by Chundng L and John Whalley. All rghts reserved. Short sectons of text, not to exceed two paragraphs, may be quoted wthout explct permsson provded that full credt, ncludng notce, s gven to the source.

2 Indrect ax Intatves and Global Rebalancng Chundng L and John Whalley NBER Workng Paper No March 212 JEL No. F1,F32,F47,H2 ABSRAC hs paper dscusses how jont cross country ndrect tax ntatves can be used to acheve global rebalancng. hs s potentally an mportant development for G2 dscussons whch thus far have centered on exchange rates as the nstruments to acheve rebalancng. We suggest that f Chna and Germany (as major surplus countres) swtch ther present VA systems from a destnaton prncple to an orgn prncple, and the US (as the major defct country) adopts a VA on a destnaton prncple VA, jontly these actons can sgnfcantly reduce the three countres jont mbalances and so contrbute to global rebalancng. We use a numercal general equlbrum model wth a monetary structure ncorporatng nsde money to capture endogenety of trade mbalances, and to also nvestgate the potental mpacts of such ntatves. hese confrm that VA structures are not only good for global rebalancng but also the changes we consder are benefcal for welfare and revenue collecton. Our research s amed to nject new deas to the present global rebalancng debate. Chundng L Insttute of World Economcs and Poltcs Chnese Academy of Socal Scences No.5 Janguomennedaje Bejng, PRC Postcode: 1732 cl428@uwo.ca John Whalley Department of Economcs Socal Scence Centre Unversty of Western Ontaro London, ON N6A 5C2 CANADA and NBER jwhalley@uwo.ca

3 1. Introducton hs paper both sets out a potental non exchange rate based global polcy ntatve nvolvng ndrect tax arrangements n the world s three largest economes (US, Germany and Chna), and, usng general equlbrum technques, evaluates ts potental mpact on mbalances both globally and for ndvdual countres. he dscusson n ths paper s relevant both for current G2 dscussons of global rebalancng and also general equlbrum modelng snce we use a modelng framework n whch the external sector mbalance for each country modeled s endogenously determned rather than beng zero or equal to some fxed exogenous amount as s more commonly done. Followng the 28 fnancal crss and the establshment of the G2 n October 28, successve G2 meetngs and declaratons have among other thngs declared the belef of G2 heads of state that global mbalances were a prmary contrbutory factor to the crss, and that efforts were needed to reduce mbalances. hese mbalances were seen as trade and current account mbalances, sharp dfferences n savngs rates across countres, and publc sector defcts and debt levels. More concretely, n Pttsburgh n September 29 the G2 commtted themselves to a Framework For Strong, Sustanable, and Balanced Growth n whch both external sector defct and surplus countres commtted to broad rangng efforts to lower ther mbalances. Snce Pttsburgh the man thrust of the G2 external sector rebalancng dscusson has centered on exchange rate polces as the mechansm to acheve rebalancng. hs n turn has lead to repeated pressure on Chna to allow apprecaton of the RMB as a 3

4 contrbuton to rebalancng. Our pont of departure s to argue that exchange rate polces are not the only avalable nstrument to acheve global rebalancng, and to focus on an alternatve nstrument, the value added tax (VA) and ts treatment of cross border transatons. Specfcally, both Chna and Germany (and the EU more broadly) operate destnaton based value added taxes under whch mports are taxed but exports leave the country tax free. Both have large trade surpluses of about 5% of GDP. Swtchng to an orgn bass whch taxes exports and allows mports tax free entry wll, gven these sgnfcant mbalances, rase taxes and effectvely also tax mbalances potentally lowerng ther sze. he long clamed neutralty of orgn/destnaton bass swtches for the VA (see Krauss and Johnson, 1974; Grossman, 198; Whalley, 1979; Genser, 1996) only holds for balanced trade, and not for today s world. In the US there s no VA, but revenue pressures gven the debt and defct stuaton could n the next few years potentally result n ts adopton. Were ths to happen, gven the large US trade defct a VA n the US ntroduced on a destnaton bass could smlarly serve to reduce the US mbalance. We also suggest that an nternatonally coordnated ndrect tax change nvolvng Chna and Germany swtchng to an orgn based VA, and the US ntroducng a destnaton based VA could potentally lead to a sgnfcant change n global external sector rebalancng. Havng set out ths possble ntatve, we then turn to an evaluaton of the potental rebalancng mpact usng general equlbrum modelng technques (see Shoven and Whalley, 1972). We use a conventonal Armngton mult-country 4

5 mult-product sngle perod trade model (see Whalley, 1985), n whch an analytcal novelty s to allow the trade mbalance for each country to be endogeneously determned. We do ths by ncorporatng nto the model both current consumpton and expected future ncremental consumpton from savngs usng an analytcal structure attrbuted to Patnkn (1947) and also adopted n Archbald and Lpsey (196) and more recently n Whalley et al (211). In an N good model, we allow preferences to be defned over both N current consumpton goods and also 1 argument representng expected future ncremental consumpton from current perod savngs. We then use a partcular functonal form that allows some countres to be savers and others borrowers, and n the Armngton trade structure ths allows for endogenous determnaton of trade mbalances. hs structure dffers from an alternaton endogenous trade mbalance smulaton for Chna recently used by L and Whalley (211), whch does not ft for the US and Germany. We calbrate ths model to 21 data on consumpton, producton, and trade, nvolvng goods n each country and use mmoble captal and labor as factor nputs. In our calculatons we ncorporate exstng ndrect tax structures by country (none federally n the US, and destnaton bass VAs n Germany and Chna). We then analyze counterfactuals nvolvng the US adoptng a destnaton based VA, and Chna and Germany adoptng an orgn based VA, and assess the mpacts on mbalances. For a combned three country tax change, mbalances fall sgnfcantly, by between 7.3% and 15.9% under dfferent US VA rates n our central case analyss. We then consder cases where these 3 countres change tax structures separately, and 5

6 wth dfferent key parameter specfcatons (elastcty and the sze of the upper bound n the nsde money formulaton). Impacts on mbalances are agan sgnfcant. he mplcaton of our analyss are that mechansms beyond exchange rates can be used to acheve ether global or country rebalancng, and one such potental mechansm s les n the ndrect tax area. 6

7 2. VAs, VA Bases and rade, and Rebalancng he value added emboded n a product s reflected n the dfference between the sale prce charged to a customer, and the cost of materals and other taxable nputs. A manufacturer remts to the government the tax rate multpled by the dfference between these two amounts, but retans taxes they had prevously pad on nputs. A VA s lke a sales tax n that ultmately only the end consumer s taxed, but dffers from a sales tax n that, wth the latter, the tax s collected and remtted to the government only once, at the pont of purchase by the end consumer. Wth the VA, remttances to the government, net of credts for taxes already pad are made every tme a busness n the supply chan sells ts products; a multstage rather than a sngle stage tax. A VA system typcally uses one of two bases n ts treatment of nternatonal trade; the destnaton prncple and the orgn prncple. he destnaton prncple mples that taxes are leved where the products are consumed for both fnal consumers and producers. Exports are exempt wth refund of nput taxes (that s, free of VA) and mports are taxed on the same bass and wth the same rates as locally produced goods. hs mples that the tax pad s determned by the tax laws n the country of consumpton and revenues accrue to the same country. he orgn prncple mples that goods are taxed where produced, regardless of where they are consumed. hs means that exportng countres tax exports on the same bass and at the same rate as domestc supples, whle mportng countres apply no VA on mports. In summary, under a destnaton prncple mports are taxed whle exports are not 7

8 taxed. Under the orgn prncple, exports are taxed whle mports are not. In the case of balanced trade, movng between these tax bases has no effect on trade flows snce t can be fnally accommodated by exchange rate or wage rate changes. In the case where, however, the trade balance s non zero and where trade s unbalanced the destnaton prncple can promote exports more than mports. he result s a larger trade surplus or smaller trade defct than would occur under an orgn bass. From the vew of rebalancng, f a country has a trade surplus, an orgn prncple n that country for the VA wll help to reduce ts surplus and ncrease revenues; and a destnaton prncple VA decease defct and ncrease revenues f the country has a trade defct. Bllon US$ year Chna Gemany Japan US Fg. 1 rade Imbalance of Some Man Countres wthn Source: WO Statstcs Database. Fgure 1 provdes data on trade mbalance countres n terms of the surplus or defct of major global tradng countres from 21 to 21. he US s man trade defct country; Chna, Germany and Japan are the man trade surplus countres n the world. Global trade mbalances worsened rapdly after 21. he fnancal crss n 28/29 relaxed the mbalance stuaton, but t worsened agan n 21. he 8

9 Japanese trade surplus decreased n the last decade and now ts mbalance s not severe. After the fnancal crss n 28, global economc recovery and rebalancng became an mportant topc of debate. Most debate focused on exchange rate adjustments to rebalance global trade. We develop the dea that exchange rate adjustment s not the only path to realzng rebalancng. VA regme swtchng can also have mpacts on rebalancng. Chna and Germany currently use a destnaton prncple VA, and the US has no VA. hus f Chna and Germany were to swtch ther VA regme from a destnaton prncple to an orgn prncple, and the US adopted a VA (whch they currently do not have) wth an orgn prncple VA, the total mbalance can be reduced wth ncreased revenues for each country (or reduced tax rates) whch also may ncrease each country s welfare. 9

10 3. Endogenous rade Imbalance Model Descrpton We use a four country general equlbrum model wth nternatonal trade n goods to whch monetary structure usng nsde money s added followng Whalley et al (211). hs allows for the endogenous determnaton of trade mbalances for trade n goods, whch s offset through nter-temporal trade across countres n money. he monetary structure bulds on Azarads (1993) where there s extensve dscusson of smple overlappng generaton models wth nsde money. In analyzng trade nteractons between monetary structure and commodty trade are needed, and hence models wth smultaneous nter-temporal and nter-commodty structure. In our general equlbrum model wth monetary structure, we assume there are two goods n each perod and allow nter-commodty trade to co-exst wthn the perod along wth trade n debt n the form of nsde money. We use a sngle perod model where ether clams on future consumpton (money holdng) or future consumpton labltes (money nsuance) enter the utlty functon as ncremental future consumpton from current perod savngs. hs s the formulaton of nsde money used by Patnkn (1947, 1971) and Archbald and Lpsey (196). hs can also be used n a mult-country model structure wth trade n both goods and nsde money. he general equlbrum model has four countres, and each country produces two goods wth two factors. hese four countres are Chna, US, Germany and the rest of the world (ROW), all trade mbalance countres n the global economy. he two goods are tradable goods and non-tradable goods. he two factors are labor and captal. On the consumpton sde, we use the Armngton assumpton of product 1

11 heterogenety across countres, and assume clams on future consumpton enter preferences and are traded between countres. Each country can thus ether ssue or buy clams on future consumpton usng current perod ncome. We use a nested CES utlty functon to capture consumpton (, N, ) [ 1 ( ) 2 ( N ) 3 ( ) ] U X X Y X X Y, country (1) Where N X denotes the consumpton of non-tradable goods n country, X denotes the consumpton of Armngton composte tradable goods n country, Y denotes the nsde money for country. Addtonally 1, 2 and 3 are share parameters and s the elastcty of consumpton substtuton. We summarze the nestng structure used n consumpton and producton n Fgure 2. he composte of tradable goods defne another nestng level reflectng the country from whch goods come. We assume ths level 2 composte consumpton s of CES form and defned as, 1 ' 1 ' ' ' ' 1 j j j X [ x ], j country (2) Where x j s the consumpton of tradable goods from country j n country. If j ths country consumes ts domestc produced tradable goods. j s the share parameter for country j ' s tradable goods consumed n country. ' s the elastcty of substtuton n level 2 n country. For a representatve consumer n country ther ncome s I, and maxmzng utlty subject to budget constrant yelds 11

12 X X I ( ) [ ( ) ( ) ( ) ] 1 1 N 1 Y 1 P 1 P 2 pc 3 pc I ( ) [ ( ) ( ) ( ) ] N 2 N 1 N 1 Y 1 pc 1 P 2 pc 3 pc I Y ( ) [ ( ) ( ) ( ) ] 3 Y 1 N 1 Y 1 pc 1 P 2 pc 3 pc (3) Where P, N pc and Y pc are separately consumpton prces of composte tradable goods, non-tradable goods and nsde money n country. For the composte tradable goods, they enter the second level and come from dfferent countres, and the country specfc demands are x ( X P ) j j ' pcj j pcj j ' (1 ) ( ) [ ( ) ] (4) where pc j s the consumpton prce of tradable goods n country from country j (produced n country j ), X P s the total expendture on tradable goods n country. he consumpton prce for the composte of tradable goods s 4 j pcj j1 P 1 ' ' (1 ) 1 [ ( ) ] (5) On the producton sde, we also assume CES technology for producton of each good n each country (Fgure 2) l l l 1 1 l l l l l l l l l 1 Q [ ( L ) (1 )( K ) ], country, l ndustry (6) where l Q s the output of the lth ndustry (ncludng tradable goods and non-tradable goods) n country, l L and K l are the labor and captal nputs, l s the scale parameter, l l s the dstrbuton parameter and s the elastcty of factor substtuton. Frst order condtons for cost mnmzaton mply the factor nput demand equatons, 12

13 K l l l L (1 ) l l l Q l w (1 ) l 1 [ [ ] (1 )] l l K w (7) l l l l l l (1 ) 1 l l K Q w L [ (1 )[ ] ] l l L (1 ) w (8) Where K w and L w are the prces of captal and labor n country. Fg. 2 Nestng Structure of Producton and Consumpton Functons Used n he Model Producton Functon (CES) Consumpton Functon (CES) radable and Non -tradable Goods Consumpton Level 1 Non-tradable Goods radable Goods Insde Money Labor Captal Level 2 Chna US Germany ROW If these countres n our model adopt a VA, each country wll receve revenues denoted as R. We assume country s ' captal and labor endowment are K and L. he total ncome of country s thus gven by w K w L R I (9) K L Budget constrants apply for each country as P X pc X pc Y I (1) N N Y Here, nsde money Y also represents country s ' trade mbalance. Y mples a trade surplus (or clams on future consumpton); Y mples a trade defct or future consumpton labltes (money ssuance), and Y mples trade balance. For trade defct countres, utlty wll decrease n nsde money snce they are ssuers. In order to capture ths, we use an upper bound Y n the utlty functon n a term [ Y Y ] followng Whalley et al (211) and assume that Y s large enough to ensure that Y Y. We use the transformaton 13 to solve the y Y Y

14 optmzaton problem, and the utlty functon and budget constrant become N N 1 (,, ) [ 1 ( ) 2 ( ) 3 ( ) ] MaxU X X Y X X y st.. P X pc X pc y I pc Y I N N Y Y * (11) he solutons to (11) are as n (3), but changng Y to y. For global trade (or money) clearance, we have Y (12) Equlbrum n the model n then gven by market clearng prces for goods and factors n each country such that Q Q N l l K L j X l l x j N K L (13) (14) Where Q and N Q are separately output of tradable goods and non-tradable goods n country. he zero proft condton must be satsfed n each ndustry n each country, such that l l K l L l p Q w K w L l, N (15) Where l p s the producer prce of goods l n country. In ths structure we can then ntroduce VA. We use t to denote the VA rate n country. A destnaton bass VA mples that 14

15 pc (1 t ) p, j mports taxed j j pc p, j exports not taxed j pc (1 t ) p, j domestc goods taxed j N N pc (1 t ) p domestc goods taxed y y pc pc nsde mo ney not taxed (16) If the VA regme s on an orgn bass, pc p, j mports not taxed j j pc (1 t ) p, j exports taxed j pc (1 t ) p, j domestc goods taxed j N N pc (1 t ) p domestc goods taxed pc y pc y nsde money not taxed (17) Usng ths general equlbrum model wth nsde money, defcts and surpluses n goods trade are endogenously determned. We can then calbrate ths structure to a base case equlbrum data set and smulate counter factual cases where Chna and Germany swtch ther present destnaton prncple VA regme to an orgn prncple regme and the US adopt a destnaton bass VA. 15

16 4. Data and Parameters Calbraton We use 21 as our base year n buldng a benchmark general equlbrum dataset for use n calbraton and counterfactual smulaton (Shoven and Whalley, 1992). here are four countres Chna, US, Germany and ROW n our model, and ROW data s obtaned by usng total world value mnus values for Chna, US and Germany. For the two tradable goods and non-tradable goods, we assume secondary ndustry (manufacturng) reflects tradable goods, and prmary and tertary ndustry (agrculture, extractve ndustres, and servces) yeld non-tradable goods. For the two factor nputs, captal and labor, we use total labor ncome (wages) to denote labor values for nputs by sector, but not labor amount. All data are n bllons of US dollars. We adjust some of the data values for mutual consstency for calbraton purposes. he data we use n model calbraton as the base case equlbrum are dsplayed n ables 1 and 2. Chnese data are from Chna data onlne. We use producton and captal values to determne labor values by resdual for Chna. US data are from the Statstcs Database of Bureau of Economc Analyss (BEA), captal and labor are from the nput-output table. German data are from OECD statstcs and the database of Germany Federal Statstcal Offce ( otal World data are from the World Economc Outlook (WEO) database of IMF, and World Development Indcators (WDI) of World Bank. From the total world data we can calculate ROW data as a resdual. rade data between each country are from the UN nteractve graphc system of nternatonal economc trends (SIGCI Plus) database (able 2). We use total export 16

17 and mport values for Chna, US and Germany to calculate these countres exports and mports wth the ROW. From producton and trade data, we can then determne each country s consumpton value. he VA rates are from Wkpeda ( nformaton data. Insde Money s gven by each country s trade mbalance and calculated from trade data n able 2. Endowment data are calculated from factor nput data. able 1 Base Year Data Used for Calbraton and Smulaton (21 Data) Item / Country Chna US Germany ROW otal Producton radable Goods Non-tradable Goods Captal Labor Insde Money Endowment radable Goods Non-tradable Goods radable Goods Non-tradable Goods Y Y y Captal Labor VA Rate 17% % 19% % Notes: (1) Unts for producton, captal, labor, nsde money and endowment are all bllon US$, and labor here denotes factor ncome (wage). (2) We use world values mnus Chna, US and Germany to generate ROW values. (3) Factor demand data reflect all factor and producton prces beng assumed to equal 1 n the base case equlbrum. Source: Chnese data from Chna data onlne; US data from the statstcs database of Bureau of Economc Analyss (BEA); German data from OECD Statstcs and database of German Federal Statstcal Offce; he World data from the World Economc Outlook database of IMF, and World Development Indcators (WDI) of the World Bank. he VA rate data from Wkpeda (he free encyclopeda). able 2 rade between Countres n 21 (Unt: bllon US$) Importer rade / Country Chna US Germany ROW Chna / Exporter US 12.7 / Germany / ROW / Note: We use each country s total export and mport value mnus values between other countres to yeld trade wth the ROW. We assume producton and factor prces equal 1. Source: UN nteractve graphc system of nternatonal economc trends (SIGCI Plus), 17

18 We choose the upper bound Y value to equal 1 n all countres. We change ths value n later senstvty analyss to check ts nfluence on counterfactual smulaton results. here are no avalable estmates of elastctes for each country on the demand and producton sde for the elastctes n our CES producton and consumpton functons. Many of the estmates of domestc and mport good substtuton elastcty are around 2 (Betna et al, 26), so we set all of these elastctes n our model to 2 (Whalley and Wang, 21). We change these elastctes later to check ther senstvty nfluence on smulaton results. able 3 reports the model parameters generated by calbraton. When used n model soluton these wll regenerate the benchmark data as an equlbrum for the model. hen, usng these parameters we smulate the effects of Chnese and German VA regme changes from the destnaton prncple to an orgn prncple and US adoptng an orgn prncple VA system. Varable/Country Share Parameters n Producton Scale Parameters n Producton able 3: Parameters Generated by Calbraton Chna US Germany ROW. N-.. N-.. N-.. N-. K L Consumpton Sde Chna US Germany ROW Chna Consumpton Share US Parameters of Level 2 Germany ROW Composte Consumpton Share N Parameters of Level 1 Insde Money Note: denotes tradable goods; N- denotes non-tradable goods. Source: Calculated usng the model structure above and calbraton method cted above by the authors. 18

19 5. Smulaton Results We next perform counter factual smulatons and report ther results. Followng the dscusson above, we explore the effects of Chnese and German VA regme swtchng from the present destnaton prncple to an orgn prncple, and of US adoptng a destnaton prncple VA. We assess VA swtchng effects on mbalances, producton, welfare and revenue both from the whole world and ndvdual countres. We generate results for four dfferent scenaros. he frst s where Chna, US and Germany make changes smultaneously, that s Chna and Germany swtch to an orgn prncple VA and US adopt a destnaton prncple VA. he second s US separately adoptng a destnaton prncple VA. he thrd s Chna s sngle swtch to an orgn prncple VA. he last s Germany s sngle swtchng to an orgn prncple VA. Chna and Germany currently have a destnaton prncple VA and we assume they use an equal tax rate swtchng method, whch means they keep ther present VA rates; just swtchng the VA bass. hs rases revenues. he swtch could be analyzed on an equal yeld bass whch lowers rates. In US, there s no VA system used at present, and we alternatvely assume ts VA rate s set at 5%, 1%, 15% and 2% n four separate cases. Snce mbalances (total export mnus total mport) can be negatve n some countres, we add all four countres absolute value of mbalances; to yeld the total world mbalance. 5.1 Smultaneous VA Changes n All hree Countres We frst consder the case where Chna and Germany swtch ther destnaton prncple VA to an orgn bass and the US adopts a destnaton prncple VA. We 19

20 compare outcomes to the base case equlbrum to get results. We assume the US VA rate s 5%, 1%, 15% and 2% alternatvely. able 4 and Fgure 3 report the results. able 4: Effects of hree Countres Smultaneously Modfyng her VA Arrangements Items / Countres Chna US Germany ROW OAL (1) Imbalance (%) US VA=5% Producton (%) Welfare (%) Revenue (%) (2) 1.4 / Imbalance (%) US VA=1% Producton (%) Welfare (%) Revenue (%) (2) / Imbalance (%) US VA=15% Producton (%) Welfare (%) Revenue (%) (2).883 / Imbalance (%) US VA=2% Producton (%) Welfare (%) Revenue (%) (2).636 / Notes: (1) otal mbalance equals the sum of each country s absolute value of mbalance; (2) he US Revenue here s the absolute value, not varaton percentage (he reason s the US VA revenue n base case s zero and we cannot calculate varaton percentage). Source: Calculated by authors. On the mbalance sde, we fnd that Chna, US, Germany and the total for the world show a decrease; only the ROW has an ncrease. hese results thus suggest that smultaneous VA changes can reduce both all ndvdual countres and the global trade mbalance, and hence contrbute to global rebalancng. Compared wth the base case data, total world mbalances separately decrease by 7.3%, 1.9%, 13.3% and 15.9% when the US VA rate used s 5%, 1%, 15% and 2%. For Chna, ts mbalance decreases are larger the hgher the US VA rate, wth changes of between -6.6% and -8.1%. For the US, ts mbalance changes ncrease as ts VA rate ncreases, wth changes of between -7.3% and -19.2%. For Germany, ts mbalance reduces substantally after the changes, decreasng from between 95.7% to 111.1% and 2

21 movng nto a defct. he reason for ths s that Germany s a hghly trade dependent country (Fgure 4) and ts exstng VA rate s hgh (19%). Its degree of dependence on trade s 76.1% n 21 n our base case data, much hgher than Chna (5.1%), US (22.3%) and the ROW(18.8%). For the ROW, mbalances wll be mproved after swtchng because of trade dverson effects. Fg. 3 Imbalance Comparsons Before and After Smultaneous VA Changes (Varatons across Assumed US VA Rate) Imbalance Comparsons Before and After Smultaneous VA Changes US VA=5% US VA=1% ,2 before after ,2 before after Chna Germany ROW OAL US Chna Germany ROW OAL US ,2 US VA=15% before after ,1,5 before after US VA=2% Chna Germany ROW OAL US Chna Germany ROW OAL US Notes: he unt s bllons US$; otal mbalance equals the sum of each country s absolute value of mbalances. Source: By authors. Fg. 4 he Degree of Dependence On rade For Indvdual Country (21 Data) rade/gdp rade/(radable Goods Producton) Chna Germany ROW US Chna Germany ROW US Source: Calculated wth the data n able 1 and able 2. 21

22 On the producton sde, changes for the world total and most ndvdual countres are negatve, because the VA regme change ncreases trade prces and hurts trade between these countres and reduces ther total producton. he reducton s, however, small. On the welfare sde, Chna and Germany beneft from these changes. her welfare ncreases by 2%-3%; US welfare falls by about -.76% and -.83% when US VA rates are 5% and 2%, because of reduced consumpton. otal world welfare mproves by between.176% and.218% after these changes. On the revenue sde, all countres gan, and especally so the US. If US uses a 5% destnaton prncple VA, ts revenue wll ncrease bllon US$; a 1% destnaton prncple VA wll collect bllon US$ and a 2% destnaton prncple VA wll collect 31.9 bllon US$. hese revenues could offset the present US government defct. We have also performed senstvty analyss for the above smulaton results. We change elastctes n producton and consumpton to separately equal.6, 1.6 and 2.6, and change the upper bound for Y to 2, 3 and 4. We then recalbrate parameters and smulate smultaneously VA change effects. Here we check the senstvty for cases where the US VA rate equals 15%. hese results are reported n able 5. We fnd that results do not change much compared wth base model smulaton results. Most changes are ncreasng n elastctes, and ncrease as the upper bound nsde money value ncrease. In general, f Chna and Germany swtch ther VA system from the present destnaton prncple to an orgn prncple and the US adopts a destnaton prncple VA, the combned effect s to reduce mbalances for the whole world and also for all 22

23 these three ndvdual countres so that t s good for global rebalancng. hese changes also mprove Chna s and Germany s welfare and ncreases all three countres revenue. able 5: Senstvty Analyss for Smultaneously Swtchng Results (US VA=15%) Items / Countres Chna US Germany ROW OAL (1) =.6 =1.6 =2.6 Y =2 Y =3 Y =4 Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2) / Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2).764 / Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2) 1.5 / Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2) / Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2) / Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2) 2.74 / Notes: (1) otal mbalance equals the sum of each country s absolute value of mbalance; (2) he US Revenue here s the absolute value, not varaton percentage (he reason s the US VA revenue n base case s zero and we cannot calculate varaton percentage). Source: Calculated by authors. 5.2 US Sngle Swtchng to ake A Destnaton Prncple VA We also smulate the effects of a US sngle adopton of a destnaton prncple VA. We agan examne cases where the US VA rate separately equals 5%, 1%, 15% and 2%. Results are presented n able 6 and Fgure 5. hese results show that the world total and all ndvdual country s mbalance 23

24 wll decrease. If US VA rate equals 1%, the whole world total mbalance wll decrease 7.47%, and Chna, US, Germany and the ROW wll separately reduce ther mbalance by.87%, 7.47%, 11.22% and 8.91%. Hence, f US adopts a destnaton prncple VA, t can not only reduce ts own trade mbalance, but also beneft global rebalancng. US producton shows no change but ts welfare and revenue wll ncrease. Chna s welfare and revenue both ncrease and producton shows nearly no change. Germany s welfare and revenue wll ncrease, but ts producton decreases a lttle. World welfare and revenue ncrease and producton declnes. able 6: Effects of US Adoptng A Destnaton Prncple VA Items / Countres Chna US Germany ROW OAL (1) Imbalance (%) VA=5% Producton (%) Welfare (%) Revenue (%) (2).25 / Imbalance (%) VA=1% Producton (%) Welfare (%) Revenue (%) (2).438 / Imbalance (%) VA=15% Producton (%) Welfare (%) Revenue (%) (2).72 / Imbalance (%) VA=2% Producton (%) Welfare (%) Revenue (%) (2).999 / Notes: (1) otal mbalance equals the sum of each country s absolute value of mbalance; (2) he US Revenue here s the absolute value, not varaton percentage (he reason s the US VA revenue n base case s zero and we cannot calculate varaton percentage). Source: Calculated by authors. We can agan change elastcty values and the upper bound on nsde money Y and do senstvty analyss. We choose elastctes to equal.6, 1.6 and 2.6, and the upper bound for nsde money to equal 2, 3 and 4 to separately check 24

25 smulaton results. able 7 gves these results. hese results do not change much compared wth base model smulaton results. Fg. 5 Imbalance Comparson Before and After US Adoptng A Destnaton Prncple VA (Unt: Bllon US$) Imbalance Comparson Before and After US akng An Destnaton Prncple VA VA=5% VA=1% Chna Germany before after Chna Germany before after ROW ROW OAL OAL US US , 1, , 1,5 VA=15% VA=2% Chna Germany before after Chna Germany before after ROW ROW OAL OAL US US , 1, , 1,5 Notes: otal mbalance equals the sum of each country s absolute value of mbalances. Source: By authors. able 7: Senstvty Analyss for US VA Adopton Results (VA=1%) Items / Countres Chna US Germany ROW OAL (1) =.6 =1.6 =2.6 Y =2 Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2) / Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2) Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2).686 / Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2) /

26 Y =3 Y =4 Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2) 1.79 / Imbalance (%) Producton (%) Welfare (%) Revenue (%) (2) / Notes: (1) otal mbalance equals the sum of each country s absolute value of mbalance; (2) he US Revenue here s the absolute value, not varaton percentage (he reason s the US VA revenue n base case s zero and we cannot calculate varaton percentage). Source: Calculated by authors. hese analyses reveal that the US adoptng a destnaton prncple VA s good for global rebalancng. In the meanwhle, t can ncrease global total and ndvdual countres welfare and ncrease ther revenues. hese effects are especally sgnfcant for the US. 5.3 Chna Sngle Swtchng VA from Destnaton to Orgn Prncple We also report results for Chna alone swtchng ts VA system from a destnaton prncple to an orgn prncple. We focus on the nfluence on mbalances, producton, welfare and revenue. able 8 and Fgure 6 show these results. As for effects on Chna, ts mbalance decreases by 15.28%, welfare ncreases by 5.2%, revenue ncreases by 3.86% and total producton decreases by about.5%. So Chna s VA system swtchng benefts Chna not only on the trade balance but also n terms of welfare and revenue. As for the nfluence on the US and Germany, both of ther mbalances ncreases, producton decreases and welfare ncreases. he ncrease n the US s mbalance s because of Chna s mproved export prce through VA ncreases on mport value, and US mportng more from Germany wth a hgher prce compared to Chna. hs suggests that the large US trade mbalance s manly caused by hgh US consumpton. 26

27 Germany s ncreased mbalance s caused by trade dverson from Chna. Although Chna s swtchng deterorates both the US and Germany mbalance t mproves ther welfare. For the world, the trade mbalance ncreases by 1.54%, total producton decreases by.84% and welfare ncreases by.13%. able 8: Effects of Chna s VA Swtchng from Destnaton to Orgn Item / Countres Chna US Germany ROW OAL (1) Imbalance (%) Producton (%) Welfare (%) Revenue (%) / / 1.93 Note: (1) otal mbalance equals the sum of each country s absolute value of mbalance. Source: Calculated by authors. Effects of Chna's Swtchng ,5 Imbalance before after Chna Germany ROW OAL US 2, 4, 6, Producton before after Chna Germany ROW OAL US 4, 8, Welfare before after Chna Germany ROW OAL US 5 1, 1,5 Revenue before after Chna Germany ROW OAL US Fg. 6 Effects of Chna s VA Swtchng to An Orgn Prncple Notes: otal mbalance equals the summaton of each country s absolute value of mbalances. Source: By authors. We perform senstvty analyss on the above results by change elastcty values and upper bound nsde money values. able 9 gves the results. We fnd that as elastctes ncrease, Chna s mbalance wll reduce more and more. he US mbalance changes from an ncrease to a decrease (when =2.6, US mbalance wll decrease.28%), the German mbalance changes from an ncrease to a decrease as well (when 27

28 =2.6, Germany mbalance wll decrease 27.23%). As upper bound nsde money ncreases, Chna mbalance further decreases (when Y =4, ts mbalance decreases by 29.5%), US mbalance further ncreases and German mbalance changes to a decrease of about 25%-28%. he man results do not change but are senstve to elastctes and the upper bound for nsde money. able 9: Senstvty Analyss for Chna VA Swtchng Results Items / Countres Chna US Germany ROW OAL (1) Imbalance (%) =.6 Producton (%) Welfare (%) Revenue (%) / / Imbalance (%) =1.6 Producton (%) Welfare (%) Revenue (%) Imbalance (%) =2.6 Producton (%) Welfare (%) Revenue (%) / / Imbalance (%) Y =2 Producton (%) Welfare (%) Revenue (%) / 1.14 / 2.78 Imbalance (%) Y =3 Producton (%) Welfare (%) Revenue (%) / / Imbalance (%) Y =4 Producton (%) Welfare (%) Revenue (%) / / Notes: (1) otal mbalance equals the sum of each country s absolute value of mbalance. Source: Calculated by authors. Chna s VA system swtchng effects reveal that t s also a good measure for Chna. It s benefcal for Chna s rebalancng and welfare. hs swtchng also mproves total world welfare but hurts global rebalancng. 5.4 Germany s Sngle VA Swtch from Destnaton to Orgn Prncple 28

29 We also explore the effects of Germany s sngle swtch of ts VA system from a destnaton prncple to an orgn prncple. able 1 and Fgure 7 report the results. able 1: Effects of Germany s VA Swtchng from Destnaton to Orgn Prncple Item / Countres Chna US Germany ROW OAL (1) Imbalance (%) Producton (%) Welfare (%) Revenue (%) / / Note: (1) otal mbalance equals the sum of each country s absolute value of mbalance. Source: Calculated by authors. Fg. 7 Effects of Germany s VA Swtchng to An Orgn Prncple Effects of Germany's VA Swtchng Imbalance Producton Chna Germany before after Chna Germany before after ROW ROW OAL OAL US US , 1,5 2, 4, 6, Welfare Revenue Chna before Chna Germany after Germany before ROW ROW after OAL OAL US US 2, 4, 6, 8, 5 1, 1,5 Notes: otal mbalance equals the summaton of each country s absolute value of mbalances. Source: By authors. For Germany, ther mbalance decreases by 63.18%, producton decreases by.98%, and welfare and revenue ncrease separately by 3.39% and 3.79%. So VA system swtchng s a good measure for Germany, and not only good for rebalancng but also benefts welfare and revenue. For Chna, ts mbalance ncreases by 2.95% caused by trade dverson effects. Producton, welfare and revenue all change negatvely. For the US, the mbalance decreases by 2.91%, producton and welfare both slghtly decrease. As for the total global changes, mbalances wll decrease 29

30 2.91%, welfare ncreases.7% and revenue ncreases 1.12%. able 11: Senstvty Analyss for Germany VA Swtchng Results Items / Countres Chna US Germany ROW OAL (1) =.6 =1.6 =2.6 Y =2 Y =3 Y =4 Imbalance (%) Producton (%) Welfare (%) Revenue (%) / 2.91 /.915 Imbalance (%) Producton (%) Welfare (%) Revenue (%) Imbalance (%) Producton (%) Welfare (%) Revenue (%) / / 1.81 Imbalance (%) Producton (%) Welfare (%) Revenue (%) / 3.79 / 1.74 Imbalance (%) Producton (%) Welfare (%) Revenue (%) / / 1.94 Imbalance (%) Producton (%) Welfare (%) Revenue (%) / / Notes: (1) otal mbalance equals the sum of each country s absolute value of mbalance. Source: Calculated by authors. We perform senstvty analyss by changng elastcty values and the upper bound nsde money value. able 11 gves the results. As elastctes ncrease, all change results ncrease, but change drectons (ncrease or decrease) do not change. In concluson, Germany VA system swtchng s good for Germany s own rebalancng and also global rebalancng, and ncreases world total and Germany welfare. 3

31 6. Conclusons We use a general equlbrum model wth monetary structure ncorporatng nsde money to yeld an endogenously determned trade surplus, and numercally calbrate t to 21 data n a 4 country sngle perod global general equlbrum model coverng the US, Germany, Chna, and the Rest of the World. We use t to capture the effects of VA regme changes (and adopton n the US) on global mbalances; an mportant ssue n current G2 dscussons. We smulate four dfferent scenaros to explore the contrbuton of these VA changes to global and ndvdual country rebalancng. hese are: (1) Chna and Germany swtchng ther present VA system from a destnaton prncple to an orgn prncple, and the US adoptng a destnaton prncple VA; (2) the US ndvdually adoptng a destnaton prncple VA; (3) Chna ndvdually swtchng to from a destnaton to an orgn prncple VA; and (4) Germany ndvdually swtchng to an orgn prncple VA. Our smulatons results suggest that any ndvdual country s VA changes wll sgnfcantly reduce world total and ndvdual country s mbalances, mprove ndvdual country s welfare and ncrease revenues. he three countres jontly makng these VA changes can sgnfcantly reduce global mbalances and mprove world welfare and most ndvdual country s welfare. In general, global mbalances wll be reduced most f Chna and Germany can swtch ther VAs from ther present destnaton prncple to an orgn prncple and the US adopts a destnaton prncple VA. Addtonally, each country s utlty wll be ncreased, and total revenue wll ncrease whch can help wth government budget defcts. 31

32 Reducng global mbalances s one of man tasks n current G2 dscussons. Most present dscussons are of exchange rate changes, whch wll produce pressure on ndvdual countres for exchange rate adjustments whch may hurt growth and ncrease unemployment. Our analyses suggest that ndrect tax ntatves such as VA prncple swtchng can be an effectve alternatve measure for achevng global rebalancng. 32

33 References Archbald, G.C. and R.G. Lpsey Monetary and Value heory: Further Comment. he Revew of economc Studes, 28(1), pp Azarads, C Intertemporal Macroeconomcs, Blackwell. Betna, V.D., R.A. McDougall and.w. Hertel. 26. GAP Verson 6 Documentaton: Chapter 2 Behavoral Parameters accessed at Genser, B A Generalzed Equvalence Property of Mxed Internatonal VA Regmes. he Scandnavan Journal of Economcs, 98(2), pp Grossman, G.M Border ax Adjustments: Do they dstort trade?. Journal of Internatonal economcs, 1, pp Krauss, M.B. and H.G. Johnson General equlbrum Analyss: A mcro-economc text. Allen & Unwn (London) Publsher. L, C and J. Whalley Rebalancng and the Chnese VA: Some Numercal Smulaton results. NBER Workng Paper w Patnkn, D Multple-Plant Frms, Cartels, and Imperfect Competton. he Quarterly Journal of Economcs, 61(2), pp Patnkn, D Insde Money, Monopoly Bank Profts, and the Real-Balance Effect: Comment. Journal of Money, Credt and Bank, 3, pp Shoven, J.B. and J. Whalley A General Equlbrum Calculaton of the Effects of Dfferental axaton of ncome from Captal n the US. Journal of Publc Economcs, 1, pp Shoven, J.B. and J. Whalley Applyng General Equlbrum. Cambrdge Unversty Press. Whalley, J Unform Domestc ax Rates, rade Dstortons and Economc Integraton. Journal of Publc Economcs, 11(2), pp Whalley, J rade Lberalzaton Among Major World radng Areas. MI Press, Cambrdge, MA. Whalley, J. and L. Wang. 21. he Impact of Renmnb Apprecaton on rade Flows and Reserve Accumulaton n a Monetary rade Model. Economc Modellng, 28, pp Whalley, J., J. Yu and S. Zhang rade Retalaton n a Monetary-rade Model. Global Economy Journal, 11(1), pp