Deficit Reductions & Multiplier Effects on Economy

Transcription

1 Deficit Reductions & Multiplie Effects on Economy Regading the Automatic Budget Enfocement Pocedues Edited Vesion Authoed by Xue Han Januay 2012

2 Table of Contents Executive Summay...1 Economic and Political Backgounds...3 Soaing Deficit and Debt Cisis... 3 Joint Select Committee on Deficit Reduction... 5 Automatic Budget Enfocement Pocedues... 6 Othe Possibilities Pending... 6 Methodology...8 Genealized Model... 8 The Investment Savings/Liquidity Pefeence Money Supply Model The IS Cuve The LM Cuve Scenaio Analysis and Adjustments Estimation of Fiscal Multiplie Application Lagging of Multiplie Effects Model by the Council of Economic Advisos Accumulation of Multiplie Effects Adjusted Pojections of Nominal GDP Adjusted Budget Outlook The Extended-Baseline Scenaio The Altenative Fiscal Scenaio Refeence Appendices... A-1 Appendix A: Data... A-1 I. Fedeal Govenment Receipt, Outlay, Deficit and Debt, A-1 II. Consumption, Disposable Income and Net Woth, A-2 III. Impots, Goss Domestic Poducts and Exchange ate of Canadian dollas... A-4 Appendix B: Regession Outputs by SAS... A-6 Appendix C: Statistical Fomulas... A-8 Glossay... G-1

3 Executive Summay This eseach epot povides compehensive analyses and conclusions fo Global Infastuctue Asset Management LLC, on thei questions egading the effects of eductions in fedeal budget deficits on the well-being of the oveall economy ove the next decade. The epot stats with cuent economic and political backgounds in the United States, employs a Keynesian appoach and the IS/LM model fo estimation of the fiscal multiplie and finally applies theoetically estimated effects of the budget cuts to adjust the pojections of goss domestic poducts. Faced by the significantly inceasing budget deficits ($1.3 tillion, 8.5% of GDP in yea 2011 as pojected) and subsequently soaing debt elative to GDP (oughly 70% by the end of yea 2011 as pojected), the Joint Select Committee on Deficit Reduction was ceated by the Budget Contol Act of Howeve, as the Joint Select Committee announced failue in late Novembe last yea, the Automatic Budget Enfocement Pocedues of $1.2 tillion in total ae expected to take place, imposing evenly distibuted cuts of oughly $110 billion pe yea fo nine fiscal yeas stating on Januay Assuming fo the Automatic Budget Enfocement Pocedues, the epot employs the IS/LM model to examine the effects of such fiscal spending cuts on each component of GDP, which enables us to adjust fo decease in govenment puchase, athe than an incease in most studies, and to account fo the inteest ate at the zeo-lowe bound. The deived fomula fo estimating the fiscal multiplie is [ ( ) ] With estimated maginal popensity to consume ( ), maginal popensity to impot ( ) and aveage tax ate ( ), the calculated multiplie is With necessay adjustments, the epot takes advantage of the model by the Council of Economic Advisos (Rome-Benstein model) fo studying the lagging effects of the fiscal multiplie. Accoding to the model, the aggegate multiplie effects of one change in govenment puchase spead ove fou yeas, with the highest in the fist yea and diminishing afte then. Applying the accumulated multiplie effects of cuts in nine consecutive yeas featued by the Automatic Budget Enfocement Pocedues and thei compound impacts to the pojections of GDP by the Congessional Budget Office, the epot aives at the adjusted pojections as shown in the table below. 1 Anothe estimated eduction of $169 billion is fom savings on debt-sevice costs. 1

4 Pojections of Nominal GDP: Befoe and Afte adjusted fo Effects of the Automatic Cuts, (Billions of U.S. dollas, unless othewise noted) CBO s Pojections Adjusted Pojections Change in GDP elated to Automatic Cuts (excluding effects of automatic cuts) (including effects of automatic cuts) ea Nominal GDP Nominal GDP Amount Amount Pecentage Nominal GDP Nominal GDP Gowth Rate (1 ea) (Compound) (Compound) Gowth Rate , % % 15, % , % % 15, % , % % 16, % , % % 16, % , % % 17, % , % % 18, % , % % 19, % , % , % 19, % , % , % 20, % , % , % 21, % , % , % 22, % , % , % 22, % , % , % 23, % As demonstated in the table above, the effects of povisions elated to the Automatic Budget Enfocement Pocedues on the oveall economy ae consideable, with compound impacts of negative $1.86 tillion, o 5.2% on aveage and 7.6% in exteme, on goss domestic poducts ove the 11-yea time hoizon fom 2013 to 2023, when compaing to pojections excluding such effects. This epot also takes a bief look at the budget outlook based on the two long-tem scenaios: the extended-baseline scenaio and the altenative fiscal scenaio. With the effects of fiscal spending cuts by the Automatic Budget Enfocement Pocedues incopoated, the pojections unde each scenaio ae depicted. 2

5 Billions of U.S. dollas Economic and Political Backgounds Soaing Deficits and Debt Cisis In the United States, the deficit ose though the 1980s and played a ole in the election of It eceded duing the Clinton administation, tuning biefly into a suplus, befoe ising again unde Geoge W. Bush. The economic meltdown in 2008 led to a spike in boowing, as tax evenues fell and the govenment stepped up stimulus spending to ty to get the economy moving. Howeve, the gowing debt also eflects an imbalance between spending and evenues that pedated the ecession. Given the county s impotance in global economy and financial makets, such issues, as well as thei possible solutions by the U.S. govenment and the effects theeafte have eceived much attention fom the wold ecently 2. As shown in Figue 1, significantly inceasing budget deficits have been ecoded by the fedeal govenment in ecent yeas as the diffeences between fedeal govenment outlay and eceipt widen. 4,000 Fedeal Govenment Receipt, Outlay & Suplus o Deficit, Fedeal Govenment Receipt Fedeal Govenment Outlay Fedeal Govenment Suplus o Deficit 3,000 2,000 1, ,000-2,000 Figue 1: Fedeal Govenment Receipt, Outlay and Suplus o Deficit, Complete data found in Appendix A-I. 3

6 Pecentage of GDP Figue 2 shows the thee budgetay components as pecentages of the economy, measued as goss domestic poducts (GDP) 3. At 8.5 pecent of GDP, the $1.3 tillion budget deficit that the Congessional Budget Office (CBO) pojects fo 2011 will be the thid-lagest shotfall in the past 65 yeas (exceeded only by the deficits of the peceding two yeas). Deficit in yea 2011 stems in pat fom the long shadow cast on the U.S. economy by the financial cisis and the ecent ecession. Although economic output began to expand again two yeas ago, the pace of the ecovey has been slow, and the economy emains in a sevee slump. Recent tumoil in financial makets in the United States and oveseas theatens to polong the slump % 25.00% Fedeal Govenment Receipt, Outlay & Suplus o Deficit As Pecentage of GDP, Receipt as Pecentage of GDP Outlay as Pecentage of GDP Suplus o Deficit as Pecentage of GDP 20.00% 15.00% 10.00% 5.00% 0.00% -5.00% % % Figue 2: Fedeal Govenment Receipt, Outlay and Suplus o Deficit, as Pecentage of GDP, Consequently, the amount of fedeal debt held by the public has suged as shown in Figue 3 on next page 4. At the end of 2008, that debt equaled 40 pecent of the nation s annual economic output (a little above the 40-yea aveage of 37 pecent). Since then, the figue has shot upwad: By the end of 2011, the fedeal debt will each oughly 70 pecent of GDP the highest pecentage since shotly afte Wold Wa II, accoding to CBO s pojection. 3 Complete data found in Appendix A-I. 4 Complete data found in Appendix A-I. 4

7 Billions of U.S. dollas 16,000.0 Nominal GDP and Fedeal Debt Held by the Public, Nominal GDP Fedeal Debt Held by the Public 14, , , , , , , Figue 3: Nominal GDP and Fedeal Debt Held by the Public, The Joint Select Committee on Deficit Reduction Facing the pofound budgetay and economic challenges, the U.S. fedeal govenment has decided to educe its budget deficit in the nea futue. The Joint Select Committee on Deficit Reduction (also called supe committee ), a 12-membe bipatisan panel, was ceated by the deal stuck by Pesident Obama and Congessional leades in late July 2011 The Budget Contol Act to allow the govenment to aise the fedeal debt ceiling enough to keep boowing into The ageement called fo at least $2 tillion in deficit eduction ove 10 yeas to offset the $2.4 tillion incease in the amount the Teasuy Depatment is authoized to boow. Out of the $2 tillion cuts, $1.2 tillion eduction in fedeal budget deficits ove 10 yeas was the stated goal of the supe committee. The act also specified automatic pocedues fo $1.2 tillion spending cuts also known as tigge cuts stating in Januay 2013, if the committee failed to agee on a plan befoe the deadline on Novembe 23 d, o if Congess did not enact its ecommendations by Decembe 23 d,

8 Automatic Budget Enfocement Pocedues ( Tigge Cuts ) The announcement of failue by the Joint Select Committee shotly befoe its deadline instated the automatic law that would impose eductions in a pocess known as sequestation. The automatic cuts, when in effect, will impose annual fiscal spending cuts of about $110 billion each yea fo the nine fiscal yeas beginning in 2013 in what is expected to emain a weakened economy. With half coming fom defense and half fom domestic pogams othe than Social Secuity, veteans benefits, militay pay and othe politically sensitive pogams, the automatic cuts bea the following featues: 1) Defense budget will be hit with an additional $600 billion of cuts in the coming decade on top of the aleady ageed to $400 billion 5. 2) Cuts to Medicae could ange fom $200 billion to $500 billion, while unde sequestation Medicaid is exempt and Medicae cuts ae limited to 2% of annual pogam spending (~$110 billion ove 9 yeas), with some exemptions 6. 3) Debt-Sevice Costs savings would sum up to an estimated eduction of $169 billion. 4) Othe Nondefense Pogams eductions anging fom 7.8 pecent (in 2013) to 5.5 pecent (in 2021) in the caps on new discetionay appopiations, esulting in outlay savings of $294 billion 7 Othe Possibilities Pending As the supe committee announced its failue to agee on a poposal of at least $1.2 tillion deficit eduction ove 10 yeas, the automatic budge enfocement pocedues, howeve, ae still not a one hunded pecent sue deal. With oughly one yea left befoe the tigge cuts take place in Januay 2013, thee ae a few othe possibilities pending that might eplace o econfigue the automatic cuts eventually. Some in Congess ae aiming at undoing the tigge cuts, but Pesident Obama theatened that he would veto any such poposals. At the meantime, Republicans ae tying to spae defense pogams by econfiguing the $1.2 tillion tigge cuts, among which $600 billion ae fom defense pogams. They believe that too much of the cuts ae weighted on the defense capabilities of the U.S. and would eally cut deeply into national defense. Anothe tendency going on within the Congess is to look backwad at poposals aleady on the table, among which the well-known Bowles-Simpson Poposal is descibed below. 5 It.aspx#page

9 Bowles-Simpson Poposal 8 is a 4 tillion, 10-yea plan offeed by Pesident Obama s National Commission on Fiscal Responsibility and Refom, also known as the Simpson-Bowles Committee. Poposal includes: 1) Riddance of all excess spending including defense, domestic pogams, entitlement spending i) Establish the Cut-and-Invest committee (Cut-and-Invest Committee to be chaged each yea with identifying 2 pecent of the discetionay budget that should be cut and identifying how to ediect half of that savings, o 1 pecent, into high-value investment) ii) Limit Emegency Spending evey fedeal agency will need to identify its excess spending iii) Requie Pesident to popose annual limits fo wa spending iv) Cut secuity and non-secuity spending (all secuity spending, which constitutes about two-thids of the discetionay budget, has one oveiding goal: to keep the nation safe. The emaining thid of the discetionay budget is dedicated to non-secuity pogams the lage aay of domestic activities, including education, housing, law enfocement, eseach, public health, cultue, povety eduction, and othe pogams) 2) Health Cae Refom elimination of administative costs associated with Medicae and Medicaid, efom of CLASS Act (ceated out of the need fo non-institutional long-tem cae, povides a voluntay long-tem insuance pogam), decease of Medicae gowth ate 3) Tax Code Refom i) Deceasing pesonal tax backets fom 5(15% 28% 31% 36% 39.6%) to 3(12% 22% 28%) ii) Single Copoate Tax backet cuently multiple backets, geneally taxed at 35% fo lage copoations must be evised to 1 backet (28%) Othe poposals that the Congess may conside besides the Bowles-Simpson include the $6 tillion, 10- yea plan poposed by Budget Committee Chaiman Paul Ryan and the nealy $4 tillion in deficit eduction ove 10 yeas poposed by a goup of senatos known as the Gang of Six. Although uncetainty still emains egading the stuctue, timing and magnitude of eductions in fedeal budget deficits ove the next 10 to 15 yeas due to the othe possibilities mentioned above, the automatic budget enfocement pocedues ae fo now the only legalized and most pobable solution fo the $1.2 tillion deficit eduction. Also, with clealy stuctued souces of eduction and even distibution of cuts among yeas, the automatic cuts povide an expectable and solid basis fo the foecasts of fedeal deficits eductions and thei impacts on the economy ove the next decade and thus will be used as a pimay assumption in this eseach. 8 f 7

10 Methodology Genealized Model In ode to study the multiplie effects of fiscal spending cuts involved in the pocedues of automatic cuts, we begin with the genealized Keynesian model to deduce the applicable fiscal multiplie. As the most impotant measue of the economy, Goss Domestic Poducts () consists of fou majo components: pesonal consumption expenditues (C), goss pivate domestic investment (I), govenment consumption expenditues and goss investment (G) and net expots of goods and sevices (NX). GDP plus tansfe payment (F) less tax evenue (R) then gives us the disposable income ( ), which is exhaustively distibuted into two categoies: consumption (C) and savings (S). (1) The amount of consumption can be futhe boken down to a constant pat as the autonomous consumption ( ) and a vaying pat as the poduct of maginal popensity to consume (c) and disposable income ( ). ( ) Similaly, the planned investment ( ) can be expessed as a constant autonomous investment ( ) and the poduct of inteest ate and a behavioal paamete (b) that measues the sensitivity of investment to inteest ate. The negative sign indicates investos favo fo low inteest ates as discount ates fo futue cash flows fom thei investments. The net expots of goods and sevices, calculated as expots less impots, can also be expessed as the poduct of maginal popensity to impot (n) and GDP subtacted fom a constant autonomous net expots ( ). The tax evenue is calculated by multiplying the aveage tax ate in U.S. and the tax base, GDP. Lastly, besides all endogenous vaiables demonstated above, thee ae two exogenous vaiables in the equation: govenment puchase (G) and tansfe payment (F) that ae completely defined by autonomous pats ( and ), espectively., 8

11 Incopoating all vaiables descibed in equations above back into Equation (1), we have the following equation, in which we futhe goup all components involving GDP () togethe and eveything else. [ ( )] ( ) ( ) (2) [( ) ] [ ( ) ] Assuming equilibium in the goods maket, the output of the economy is equal to the total planned expenditue ( ). Combining Equation (2) and (3) leads us to Equation (4). (3) [ ( )] ( ) ( ) (4) [( ) ] [ ( ) ] ( ) [ ( ) ] whee the total autonomous planned expenditue ( ), is composed of all constant tems in the equation and thus is a constant as well. Equation (4) can be then illustated as a staight line in the diagam of goods maket shown in Figue 4, with a vetical intecept at ( ) and a slope of [ ( ) ]. The 45 degee line epesents the equilibium in the goods maket and its intecept with line indicates the equilibium level of GDP ( ). Goods Maket (A p b ) [c( t) n] A p b slope c( t) n Figue 4: Goods Maket, equilibium at intecept 9

12 In Figue 4, the line that featues Equation (4) has a slope smalle than one (flatte than the 45 degee line), which aises fom the facts that both maginal popensity to consume (c) and maginal popensity to impot (n) in the slope coefficient ae numbes geate than 0 and less than 1; togethe with a positive aveage tax ate (t), the slope coefficient [ ( ) ] must be less than 1. This esult is also the undelying eason fo the multiplying effect of fiscal spending to be discussed in details late. As the fedeal govenment implements fiscal policies to incease o decease govenment spending, we obseve a diect change in govenment puchase ( ), contibuting to the change in GDP ( ) as a component of GDP. In most cases, howeve, we obseve fiscal multiplies not exactly equal to one, implying that besides the diect change in govenment puchase, thee ae othe factos affecting the total change in GDP ( ) given a stimulus o cut in fiscal spending. Specifically, these factos ae the impacts of the changes in the thee othe components of GDP: Consumption (C), Investment (I) and Net Expots (NX). Befoe we go ahead estimating the size of the multiplie, it is vey necessay to examine how the effects of a change in govenment puchase flow though and inteact with each component in GDP. This appoach enables us to caefully examine cetain conditions and appopiately adjust them to eal wold situations; it is also helpful in studying the lagging effects of a fiscal multiplie. 10

13 The Investment Savings/Liquidity Pefeence Money Supply Model To examine the changes in each of the fou components, we employ the IS/LM model (Investment Saving/Liquidity Pefeence Money Supply), which demonstates the elationship between inteest ates and eal output in both the goods maket and the money maket. The IS Cuve Fistly, Figue 5 on the left illustates how the IS cuve is constucted. The IS cuve, epesenting a locus of all equilibia between total spending and total output of the economy, is a downwad-sloping line in the demand side with inteest ate () on the vetical axis and GDP () on the hoizontal axis. When inteest ate inceases fom to, we obseve a downwad paallel shift of line to line as the vetical intecept deceases fom to. Intuitively, when inteest ate inceases, total business investment shinks, leading to a smalle GDP and hence a lowe level of expenditue in the goods maket. Its intecept with the 45 degee line moves fom point A to point B, esulting in a lowe GDP ( ). This change is eflected in the demand side as a move along the IS cuve fom the oiginal equilibium at A ( ) to the equilibium aftewads at B ( ). Point and in these gaphs indicate non-equilibium situations whee planned expenditue is eithe geate o less than total output. A p A p What happens to the IS cuve in face of a change in govenment puchase then? Conside a stimulus in govenment puchase ( ) fo now, which is the most common and conventional scenaio in studies of multiplie. As illustated in Figue 6, we obseve an upwad paallel shift of line to line as the vetical intecept inceases fom to. As GDP inceases to a highe level ( ) while inteest ate holds constant at, the IS cuve on the demand side shifts to the ight fom to. Goods Maket ( ) A b α β (E b B p < ) B A p G b G A p b A Goods Maket Demand Side Demand Side B β α A IS ( ) A B IS ( ) IS ( ) Figue 5: A move along the IS cuve Figue 6: A shift of the IS cuve 11

14 The LM Cuve As illustated in Figue 7, the LM cuve is constucted in a vey simila way as that of the IS cuve. In the money maket diagam with inteest ate () on the vetical axis and liquidity ( ) on the hoizontal axis, the downwad-sloping liquidity pefeence function ( ) epesents willingness to hold cash balances instead of secuities, o in othe wods, quantity of cash balances demanded. The money supply function ( ) is also plotted in the same diagam as a vetical line, because it is detemined by the cental bank s decision and thus featues pefect inelasticity with espect to inteest ate and othe factos. The intecept of the liquidity pefeence function and the money supply function epesents a money maket equilibium, the collection of which foms the LM cuve in the diagam of the demand side. When GDP inceases fom to, the liquidity pefeence function makes a paallel shift to the ight, esulting in a highe inteest ate ( ) at the same level of cash balance. Such change is eflected in the demand side of Figue 7 as a move along the LM cuve fom A to B. When monetay policy makes take steps, fo example, to push inteest ates down, a ightwad shift of the money supply function is obseved as money supply inceases. The outcome is a lowe inteest ate ( ) with the size of total output ( ) unchanged. Reflected on the diagam of the demand side in Figue 8 is a shift of the LM cuve to. Money Maket Demand Side M s P LM(L M s P ) β B (L < Ms P ) (L M s P ) A α L( ) L( ) M P Figue 7: A move along the LM cuve β A B α Money Maket M s P M s P Demand Side LM LM A B A B L( ) M P Figue 8: A shift of the LM cuve 12

15 Scenaio Analysis and Adjustments With necessay knowledge of the IS cuve and LM cuve, we now connect these things togethe and build the IS/LM model shown in the following figues. We fistly conside the most common and simple scenaio, whee a fiscal policy featuing an incease in govenment puchase ( ) is implemented and monetay policy makes do not step in. Unde such assumptions, we see an upwad shift of line to line, esulting in a highe GDP at and a ightwad shift of IS cuve. Meantime in the money maket, with no change in money supply, the LM cuve stays at the same place and intecepts with the new IS cuve at a new equilibium ( ). The highe inteest ate coesponds to the shift of the liquidity function fom ( ) to ( ) in the money maket. Such outcome of a final GDP at a lowe level than the tageted is a esult of the cowding out of investment, which efes to the negative impact on investment due to highe inteest ates that an incease in fiscal spending can cause. Goods Maket E p A p G b A p b G Demand Side LM(L M s P ) Money Maket M s P IS ( ) IS ( ) Figue 9: Scenaio 1, Incease in Govenment Puchase, No Monetay Policies L( ) L( ) M P 13

16 In the next scenaio, we conside monetay policy makes who step in to complement the expansionay fiscal policies by avoiding the cowding out of investment using monetay policies. While govenment puchase inceases and IS cuve shifts to the ight, monetay policy makes incease money supply and shift the money supply function ightwad, esulting in a shift of the LM cuve. The new IS cuve and new LM cuve intecepted at a new equilibium ( ). With inteest ate held constant, thee is no hut on investment and thus GDP aives at the tageted level of. Goods Maket A p G b A p b G Demand Side LM Money Maket M s P M s P LM IS IS L( ) L( ) Figue 10: Scenaio 2, Incease in Govenment Puchase, Complementay Monetay Policies M P 14

17 In face of the automatic cuts in the coming decade, on the opposite of a fiscal stimulus, govenment puchase will expeience a decease ( < ). Illustated in Figue 11, as line makes a downwad shift to line, IS cuve shifts back to the left and intecepts with LM cuve at a lowe inteest ate. Assuming monetay policy makes take no actions, the lowe inteest ate benefits investment and hence pushes line back up to line, aiving at the final equilibium with GDP at and inteest ate at. The stoy told hee is an invese of the cowding out of investment in the case of a contactionay distubance. Goods Maket A p b A p G b G < E p Demand Side LM(L M s P ) Money Maket M s P IS ( ) IS ( ) Figue 11: Scenaio 3, Decease in Govenment Puchase, No Monetay Policies L( ) L( ) M P 15

18 Thee is anothe impotant adjustment to be made in ode to accuately descibe and analyze the multiplie effect in ou cuent eal wold situation. In scenaio 3 analyzed above, one outcome of the cuts in govenment spending is a lowe inteest ate. Howeve, with the 3-month T-bill at an inteest ate of 0.03% 9, the inteest ate that Fedeal Reseve has contol on is aleady at the so-called zeo-lowe bound. With no much oom left fo the inteest ate to go down, the Fedeal Reseve would have to take steps to hold the inteest ate at its cuent level in face of the automatic cuts on fiscal spending. The last scenaio and also the one that fits ou eal wold situation best is pesented below in Figue 12, with a negative change in govenment spending and a coesponding decease in money supply to hold inteest ate constant. In this case, the new IS and LM cuve intecept at equilibium ( ). The absolute multiplying effect of govenment spending cuts is geate than that in the last scenaio with an invese cowding out of investment. Goods Maket A p b A p G b G < Demand Side Money Maket M s P M s P LM LM IS IS Figue 12: Scenaio 4, Decease in Govenment Puchase, Complementay Monetay Policies L( ) L( ) M P 9 Infomation updated on Januay 13th,

19 Estimation of Fiscal Multiplie With delibeate analysis of the eal wold situation we face using the IS/LM model with adjustment to account fo the inteest ate at the zeo-lowe bound, we may now simplify the genealized model fo estimating multiplie. As inteest ates () ae assumed to hold at a constant level, the component of planned investment ( ) can be then expessed completely by a constant. Fo the pupose of easy undestanding, let the planned investment ( ) equal to the autonomous investment ( ), whee incopoates the poduct of inteest ate and the behavioal paamete (b) as well. now Then ou genealized model can be simplified as below to a line with vetical intecept of of [ ( ) ]. and a slope [ ( )] ( ) (5) [ ] [ ( ) ] [ ( ) ] Demonstated below in Figue 13 is the case of multiplie effects with the automatic cuts and inteest ates at zeo-lowe bound taken into account. As govenment puchase deceases by the amount of, total planned autonomous expenditue ( ) deceases by the same amount, shifting line downwad to aive at the new equilibium with output at. Boken down in the diagam, the decease in GDP ( ) comes fom two souces: the diect decease in govenment puchase ( ) and an indiect decease [ ( ) ], which is eally the effects on consumption and net expots ( and ). [ ( ) ] ( ) Since investment is not affected as inteest ates ae held constant, we have accounted fo the effects on all fou components in such case. Goods Maket A p [c( t) n] slope [c( t) n] A p G (A p G) [c( t) n] A p G < G < [c( t) n] c( t) n C NX < Figue 13: Goods Maket, Decease in Govenment Puchase, Stable Inteest Rates 17

20 Afte caefully examine and adjust fo all aspects, we can now safely deive the fiscal multiplie as following. By definition, the fiscal multiplie is measued as the change in GDP pe unit change in govenment puchase, that is. { [ ( ) ]} [ ( ) ] [ ( ) ] [ ( ) ] To estimate the size of the multiplie, we need to estimate the thee indicatos involved in its calculation fist: maginal popensity to consume (c), maginal popensity to impot (n) and aveage tax ate (t). Maginal Popensity to Consume The maginal popensity to consume, by its definition, seves as the slope coefficient of the equation below whee Autonomous Consumption ( ) is the vetical intecept. To estimate the slope coefficient, we can constuct the equation above as a egession with consumption (C) as dependent vaiable and disposable income ( ) as independent vaiable. Howeve, since both vaiables featue non-stationay time seies data, such a egession would suffe fom autocoelation poblem and hence yield biased estimates fo the coefficients. The solution fo autocoelation is the basic and widely-used method in econometics Fist Diffeence Method, whee we use the fist diffeence of each non-stationay time seies vaiable in the egession instead of itself 10. The egession then becomes ( ) Using the fist diffeences of time seies data on consumption and disposable income fom 1962 to , the egession is estimated as ( ) ( ) ( ) 10 Fist Diffeence Method also featues egession though the oigin, i.e. without a constant tem. 11 Complete data found in Appendix A-II. 18

21 Anothe impotant vaiable that explains vaiations in consumption is the total wealth, measued as net woth (NW). The level of consumption not only has things to do with the size of the economic output, but also is affected by the accumulated wealth, and leaving out such vaiable might lead to biased estimates in egession. Adding the fist diffeence of net woth as anothe independent vaiable, ou egession model now becomes ( ) ( ) Using the fist diffeences of time seies data on consumption, disposable income and net woth fom 1962 to , we estimate the above egession as shown below. ( ) ( ) ( ) ( ) ( ) An incemental F-test of and an Adjusted impoved fom to both indicate that the newly added independent vaiable of net woth has impoved the model and esulted estimates close to thei tue values. Also, with both coefficients estimated in the egession statistically significant 14 and a global F-test of , the oveall model has high explanatoy powe and yields elatively accuate estimations. Theefoe, we adopt the coefficient estimate as ou maginal popensity to consume (c). Maginal Popensity to Impot Similaly with maginal popensity to consume, maginal popensity to impot seves as the slope coefficient of the equation below. whee the autonomous net expots ( ) accounts fo all the expot and the autonomous potion of impots, togethe as the vetical intecept. By subtacting the expots fom both sides of the equation, we have an equation with only impots and GDP involved. Again, we utilize the Fist Diffeence Method to avoid autocoelation poblem and tansfom the equation as following ( ) 12 Complete data found in Appendix A-II. 13 The citical value of an incemental F-test with degees of feedom m=1 and n=46 at 99 pecent confidence level is Fo a coefficient to be statistically significant, its t-scoe should exceeds the citical value. The citical value of a t- test with 46 degees of feedom at 99 pecent confidence level is The citical value of a global F-test with degees of feedom m=1 and n=46 at 99 pecent confidence level is

22 With time seies data on impots and nominal GDP fom 1962 to , we estimate the egession above ( ) ( ) ( ) Howeve, anothe vaiable that plays an impotant ole in detemining the level of impots being left out is the exchange ate (e) of majo cuencies with the U.S. dolla. In ode to account fo the impacts of exchange ate without ovecomplicating the model with too many vaiables, hee we choose the use the exchange ate of Canadian dolla with the U.S. dolla, fo that Canada is the lagest tading patne of the U.S.. Now the egession model becomes ( ) ( ) Using time seies data on impots, GDP and exchange ate of Canadian dolla fom 1962 to , the egession is estimated as ( ) ( ) ( ) ( ) ( ) The impovement in Adjusted fom to , as well as an incemental F-test of implies that the newly added independent vaiable of exchange ate has impoved the model and esulted in estimates close to thei tue values. Also, with both coefficients estimated in the egession statistically significant 19 and a global F-test of , the oveall model has high explanatoy powe and yields elatively accuate estimations. Theefoe, we adopt the coefficient estimate as ou maginal popensity to impot (n). Aveage Tax Rate Lastly, we estimate the aveage tax ate using the most ecent numbes in By its definition, the aveage tax ate is calculated as dividing total tax evenue (R) by the tax base, which is the nominal GDP (). 16 Complete data found in Appendix A-III. 17 Complete data found in Appendix A-III. 18 The citical value of a global F-test with degees of feedom m=1 and n=46 at 90 pecent confidence level is Fo a coefficient to be statistically significant, its t-scoe should exceeds the citical value. The citical value of a t- test with 46 degees of feedom at 95 pecent confidence level is The citical value of a global F-test with degees of feedom m=1 and n=46 at 99 pecent confidence level is

23 With all thee indicatos involved estimated above, the fiscal multiplie is calculated as [ ( ) ] [ ( ) ] Out of this multiplie of 1.88, the effect fom the indiect change in consumption is ( ) ( ) ( ) And the effect fom the indiect change in net expots is This estimate of the fiscal multiplie is widely ageed by many economists and eseaches in thei wok. Fo example, accoding to Ramey (2011), the aggegate multiplie of a govenment puchase should be geate than 0.8 and less than 1.5, with the acceptable ange fom 0.5 to 2.0 fo adjustments egading states of the economy and movements of inteest ates. In ou case, the multiplie fo estimation is unde the cicumstance of a ecession, which, accoding to multiple studies, would esult in a highe multiplie. Anothe found well admitted by many eseaches is that with an inteest ate at zeo-lowe bound, fiscal multiplie tends to be highe than its value othewise. These two easons should explain the small spead between ou estimate of 1.88 and the most commonly obseved value of 1.5 fo a multiplie. Anothe study that concluded vey simila esults with ou estimation is by Rome and Benstein (2009). With the vey same extenal conditions and simila estimation of the multiplie, this quately-basis model by Rome and Benstein will be discussed in details next to examine the lagging effects. 21

24 Application Lagging of Multiplie Effects Model by the Council of Economic Advisos (The Rome-Benstein Model) The multiplie effects of a change in govenment spending ae execised neithe immediately afte the change no in a single peiod of time. Instead, the effects usually stat to show shotly afte the change is implemented and last fo a quite long time as funds flow though consumption and net expots. In ode the examine the lagging effect of the multiplie, we adopt the model povided by the Council of Economic Advisos (CEA) in thei estimation of job ceation fom the Ameican Recovey and Reinvestment Act (ARRA) of This model was oiginally developed by aveaging the multiplies fom a leading pivate foecasting fim and the Fedeal Reseve s FRB/US model by Chistina Rome and Jaed Benstein, the fome of which was the Chai of the Council of Economic Advisos in the Obama administation fom Januay 2009 to Septembe On a quately basis, this model caefully examines the lagging effect of a multiplie ove a peiod of 16 quates with the aggegate value ising to the highest at 1.57 by the 8 th quate as epoduced in Table 1. Table 1: Multiplie effects of a pemanent stimulus of 1 unit in govenment puchase on GDP Quate Size of Multiplie Fo the pupose of ou eseach, it is necessay to make the following adjustments to the CEA model pesented above. Fistly, the model should be conveted to an annual basis fo studying the effects of yealy automatic cuts. Secondly, the aggegate multiplie of 1.57 in the CEA model should be adjusted on a popotional basis to meet ou estimation of The two adjustments ae shown in Table 2 on the next page

25 Table 2: Adjustments fom CEA model to Annual Basis and Aggegate Multiplie 1.88 Quately Annual Aveage Annual Aveage Adjusted Annual Quate ea Aggegate Effects Effects (Aggegate) Effects (1 ea) Effects (1 ea) Accumulation of Multiplie Effects In ou cuent situation, instead of one change in govenment puchase, the automatic cuts include evenly distibuted cuts in nine consecutive yeas. Thus, given the adjusted annual multiplie effects, we now need to conside the accumulation of multiplie effects fom changes in diffeent yeas in evey single yea duing the time hoizon of the automatic cuts. Fo instance, in the fist yea of implementation of the automatic cuts, only ea 1 s multiplying effects in the fist yea is execised, that is 1.52; in the second yea, the effect of the ea 1 s cuts in the second yea and that of ea 2 s cuts in the fist yea togethe give the accumulated multiplie effects in that yea. The exact pocedues ae shown in Table 3. Table 3: Accumulation of Multiplie Effects of Nine Consecutive Cuts in each yea Automatic Cuts Accumulated ea Multiplie ea 1 ea 2 ea 3 ea 4 ea 5 ea 6 ea 7 ea 8 ea 9 Effects

26 Adjusted Pojections of Nominal GDP As discussed in the fist pat of the eseach, the Automatic Budget Enfocement Pocedues featue evenly distibuted annual cuts of $110 billion fo 9 consecutive yeas stating Applying the accumulated multiplie effects calculated above, we have the impacts of the automatic cuts on GDP in dolla amounts fo each yea, which futhe give us the compound impacts ove eleven yeas. The esults ae shown in Table 4. Table 4: Multiplie Effects of the Automatic Budget Enfocement Pocedues on GDP (Billions of U.S. dollas, except fo size of multiplies) ea Amount of Automatic Accumulated Amount of Change Amount of Change Cuts (Billions of $) Multiplie Effects in GDP (1 ea) in GDP (Compound) Let s emind ouselves about the decomposition of the multiplie effects into souces fom diect change in govenment puchase and indiect changes in consumption and net expots. To have a close look at the impacts of the automatic cuts on the economy, we can beak down to see the multiplies and amounts of changes in consumption and net expots in each yea and compound, which ae shown in Table 5. Table 5: Decomposition of Multiplie Effects into Consumption and Net Expots ea Accumulated Multiplie Amount of Change in Amount of Change in Amount of Effects on (1 ea) (Compound) Automatic Cuts Consumption Net Expots Consumption Net Expots Consumption Net Expots

27 Based on the Congessional Budget Office (CBO) s pojections of nominal GDP excluding effects of povisions elated to automatic cuts fo yeas though 2023, we can poject the size of the nominal GDP with the multiplie effects of such povisions included as shown in Table 6. The hut caused by the automatic cuts on economy can be measued by the dolla amount o pecentage decease in GDP afte including thei effects. Also, gowth ate of GDP in each case is calculated and shown, espectively. Table 6: Pojections of Nominal GDP: Befoe and Afte adjusted fo Effects of the Automatic Cuts, (Billions of U.S. dollas, unless othewise noted) CBO s Pojections Adjusted Pojections Change in GDP elated to Automatic Cuts (excluding effects of automatic cuts) (including effects of automatic cuts) ea Nominal GDP Nominal GDP Amount Amount Pecentage Nominal GDP Nominal GDP Gowth Rate (1 ea) (Compound) (Compound) Gowth Rate , % % 15, % , % % 15, % , % % 16, % , % % 16, % , % % 17, % , % % 18, % , % % 19, % , % , % 19, % , % , % 20, % , % , % 21, % , % , % 22, % , % , % 22, % , % , % 23, % 25

28 Billions of U.S. dollas The pojections of GDP both befoe and afte adjusted fo effects elated to povisions of the automatic cuts ae also gaphed in Figue 14 shown below. As we can tell fom the data and the gaph, the negative impacts on the oveall economy caused by the automatic cuts of $1.2 tillion ae consideable, with a compound amount of negative $1.86 tillion, o 5.2% on aveage and 7.6% in exteme, on goss domestic poducts ove the 11-yea time hoizon fom 2013 to 2023, when compaing to pojections excluding such effects. Pojections of Nominal GDP: Befoe and Afte Adjustments fo Effects of Automatic Cuts CBO's Pojections of GDP Actual Pojected Pojections of GDP Adjusted fo Effects of Automatic Cuts Automatic Cuts Stat Figue 14: Pojections of Nominal GDP Befoe and Afte Adjustments fo effects of Automatic Cuts 26

29 Adjusted Budget Outlook Equipped with ideas about how the economy would be impacted assuming the automatic cuts to take place, now let s take a bief look at budgetay side of the stoy, moe specifically the tends of govenment deficit and debt with the changes fom the automatic cuts taken into account. In thei 2011 Long-Tem Budget Outlook, the Congessional Budget Office (CBO) povided two scenaios fo the analysis of the issue: the extended-baseline scenaio and the altenative fiscal scenaio. The Extended-Baseline Scenaio The extended-baseline scenaio adhees closely to cuent law, following CBO s 10-yea baseline budget pojections though 2021 and then extending the baseline concept fo the est of the long-tem pojection peiod. Unde this scenaio, the expiation of the tax cuts enacted since 2001 and most ecent extended in 2010, the gowing each of the altenative minimum tax, the tax povisions of the ecent health cae legislation, and the way in which the tax system inteacts with economic gowth would esult in steadily highe evenues elative to GDP. At the same time, govenment spending on eveything othe than the majo mandatoy health cae pogams, Social Secuity, and inteest on fedeal debt activities such as national defense and a wide vaiety of domestic pogams would decline to the lowest pecentage of GDP since befoe Wold Wa II. The significant incease in evenues and decease in the elative magnitude of othe spending would offset much though not all of the ise in spending on health cae pogams and Social Secuity. As a esult, debt would incease slowly fom its aleady high levels elative to GDP, as would the equied inteest payments on that debt. The detailed pojections ae pesented in Table 7. Table 7: CBO s Budget Pojections unde the Extended-Baseline Scenaio, (Billions of U.S. dollas, unless othewise noted) ea Pojections Excluding Effects Debt Held by the Public Effects of Total of Automatic Cuts at the End of the ea Automatic Cuts Deficit Revenues Outlays Deficit Amount Pecentage of GDP ,163 3,456-1, ,294 9, ,314 3,597-1, ,284 10, ,635 3, , ,069 3, , ,423 3, , ,665 3, , ,847 4, , ,087 4, , ,286 4, , ,508 4, , ,731 5, , ,969 5, ,

30 The Altenative Fiscal Scenaio The budget outlook is much bleake unde the altenative fiscal scenaio, which incopoates seveal changes to cuent law that ae widely expected to occu o that would modify some povisions of law that might be difficult to sustain fo a long peiod. Most impotant ae the assumptions about evenues: that the tax cuts enacted since 2001 and extended most ecently in 2010 will be extended; that the each of the altenative minimum tax will be estained to stay close to its histoical extent; and that ove the long un, tax law will evolve futhe so that evenues emain nea thei histoical aveage of 18 pecent of GDP. This scenaio also incopoates assumptions that Medicae s payment ates fo physicians will emain at cuent levels (athe than declining by about a thid, as unde cuent law) and that some policies enacted in the Mach 2010 health cae legislation to estain gowth in fedeal health cae spending will not continue in effect afte In addition, the altenative scenaio includes an assumption that spending on activities othe than the majo mandatoy health cae pogams, Social Secuity, and inteest on the debt will not fall quite as low as unde the extended-baseline scenaio, although it will still fall to its lowest level (elative to GDP) since befoe Wold Wa II. Unde those policies, fedeal debt would gow much moe apidly than unde the extended-baseline scenaio. With significantly lowe evenues and highe outlays, debt held by the public would exceed 100 pecent of GDP by 2021.Afte that, the gowing imbalance between evenues and spending, combined with spialing inteest payments, would swiftly push debt to highe and highe levels. The detailed pojections ae epoduced and adjusted as below. Table 8: CBO s Budget Pojections unde the Altenative Fiscal Scenaio, (Billions of U.S. dollas, unless othewise noted) ea Pojections Excluding Effects Debt Held by the Public Effects of Total of Automatic Cuts at the End of the ea Automatic Cuts Deficit Revenues Outlays Deficit Amount Pecentage of GDP ,163 3,456-1, ,294 9, ,234 3,638-1, ,404 10, ,537 3,665-1, ,128 11, ,751 3,754-1, , ,970 3, , ,191 4,243-1, , ,363 4,567-1, ,080 15, ,605 4,847-1, ,110 16, ,792 5,111-1, ,181 17, ,986 5,475-1, ,343 18, ,183 5,828-1, ,492 20, ,385 6,172-1, ,626 22,

31 Pecentage of GDP The gowth of debt elative to GDP in each scenaio is also illustated in Figue 15 shown below. 200% 180% 160% 140% Fedeal Debt Held by the Public unde CBO's Long Tem Budget Scenaios Actual Pojected Altenative Fiscal Scenaio 120% 100% 80% 60% Extended-Baseline Scenaio 40% 20% 0% Figue 15: Fedeal Debt Held by the Public Relative to GDP unde CBO s Long-tem Budget Scenaios 29

32 Refeences Ramey, Valeie A Can Govenment Puchases Stimulate the Economy? Jounal of Economic Liteatue 2011, 49:3, Auebach, Alan, and uiy Goodnichenko. Fothcoming. Measuing the Output Responses to Fiscal Policy. Ameican Economic Jounal: Economic Policy Rome, Chistina D., and Jaed Benstein The Job Impact of the Ameican Recovey and Reinvestment Plan. Executive Office of the Pesident, Council of Economic Advisos Estimates of Job Ceation Fom The Ameican Recovey and Reinvestment Act of Congessional Budget Office CBO s 2011 Long-Tem Budget Outlook. Congessional Budget Office The Budget and Economic Outlook: An Update. 30

33 Appendices Appendix A: Data I. Fedeal Govenment Receipt, Outlay, Deficit and Debt, (Billions of U.S. dollas, unless othewise noted) Receipt Outlay Suplus o Deficit (-) Debt Held by the Public ea Pecentage Pecentage Pecentage Pecentage Amount Amount Amount Amount of GDP of GDP of GDP of GDP % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % 1, % % % % 1, % % % % 1, % % % % 1, % % 1, % % 1, % % 1, % % 2, % % 1, % % 2, % , % 1, % % 2, % , % 1, % % 2, % , % 1, % % 3, % A-1