When is Austerity Ineffective?

Marattin, Luigi Working Paper When is Austerity Ineffective? Quaderni - Working Paper DSE, No. 880 Provided in Cooperation with: University of Bologna, Department of Economics Suggested Citation: Marattin, Luigi (2013) : When is Austerity Ineffective?, Quaderni - Working Paper DSE, No. 880, Alma Mater Studiorum - Università di Bologna, Dipartimento di Scienze Economiche (DSE), Bologna, https://doi.org/10.6092/unibo/amsacta/3679 This Version is available at: https://hdl.handle.net/10419/159719 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc/3.0/ When is Austerity Ineffective? Luigi Marattin Quaderni - Working Paper DSE N°880 When is Austerity Ine¤ective ? Luigi Marattin May 14, 2013 Abstract This paper o¤ers a formal analysis of the relationship between changes in government primary balance and debt-to-GDP ratio. it establishes the conditions under which a …scal consolidation increases - instead of decreasing - the stock of government liabilities relative to aggregate output. A crucial role is played by the relationship between the elasticities of average cost of debt and nominal output to primary balance: while the former depends on debt maturity and risk premia dynamics, the latter relates to the well-known controversy on the size of government spending multipliers. The paper shows an application to the ongoing …scal consolidation process in the Eurozone. JEL Classi…cation: E62, H62 Keywords: debt-to-GDP ratio, debt sustainability, government budget. I would like to thank Roberto Golinelli and Arsen Palestini for useful hints and suggestions. Usual disclaimer applies. 1 1 Introduction As most economies have been implementing severe …scal consolidation processes, the debate is open on the e¤ectiveness of what has been called "economic austerity" (Gros 2011,Blanchard and Leigh 2013, Mauro et al 2013). If the reduction of debt-to-GDP ratios is believed to be the most appropriate measure of the e¤ectiveness of …scal consolidations processes, it is necessary to acknowledge the presence of several ways in which changes in the …scal position can actually achieve the objective, at least in the short-run. This short note outlines a simple theoretical framework in which such a transmission mechanism can be properly analyzed. Using the arithmetic of government budget constraint, this note identi…es and discusses the conditions under which a change in primary balance causes a change of the same sign in the debt-to-GDP ratio. We employ this condition as a measure of the e¤ectiveness of austerity. Our starting point is that there are three forces acting on the transmission chain from primary budget changes to debt-to-GDP dynamics. One is direct: ceteris paribus, an increase/decrease in the government net absoption of resources causes an increase/decrease of the stock of nominal debt. The other two are indirect, and work in opposite ways: an increase (decrease) in primary de…cit increases (decreases) nominal growth in the short-run and therefore reduces (increases) debt-to-GDP ratio. At the same time, by modifying risk premia on the existing and new debt, increases (decreases) the average cost of debt and therefore increases (decreases) debt-to-GDP . We consider the government primary balance as the primitive …scal policy instrument, disregarding here the disaggregation between revenue and expenditure components of the budget. Our results show that a given primary budget reduction is successful in causing an actual decrease of debt-to-GDP ratio only if elasticities of debt’s average cost and nominal output with respect to primary de…cit satisfy certain conditions. The natural heterogeneity of those conditions across di¤erent economies (according to the composition of the total budget) can have a coordinated …scal policy move result in opposite e¤ects, even with similar values for elasticities. The rest of the paper is organized as follows. Section 2 outlines the setting and identi…es the conditions, providing also a graphical representation and an application to Euro-area economies. Section 3 discusses analytically the results and section 4 o¤ers some concluding remarks. 2 Conditions for e¤ectiveness of austerity This section identi…es and discusses the conditions connecting changes in the policy instrument to their e¤ects on the debt-to-GDP ratio. We also present an application to major EMU economies. The starting point is the simple nominal government ‡ow budget constraint: 2 Bt=#t(Dpr t)Bt1+Dpr t(1) where Btis the end-of-the-period stock of government nominal liabilities, #tis the average cost of debt (obtained by dividing interest payments by total stock of gross government liabilities) and Dpr tis the primary budget de…cit1. We assume that #tis a function of Dpr t, as changes in primary budget a¤ect risk premia on new debt and therefore its average cost, even though the last step crucially depends on debt’s maturity structure. In order to express (1) in terms of debt-to-GDP dynamics, we divide it by nominal GDP (Yt): Bt Yt =#t(Dpr t)Bt1 Yt(Dpr t)+Dpr t Yt(Dpr t)(2) The above expression is the familiar law of motion of the stock of government liabilities relative to aggregate output (Marattin and Marzo 2010). Assuming the policy instrument to be the primary de…cit Dpr t(>0);we analyze how Bt Yt is a¤ected by policy changes. Equation (2) explicitly takes into account that changes in primary de…cit contemporaneously a¤ect the level of income. Partial derivative of (2) with respect to Dpr treads: @Bt Yt @Dpr t =Bt1 Yt@#t @Dpr t #t Dpr t "DY t1 Yt"DY t1 with "DY tbeing the elasticity of nominal output with respect to primary balance: "DY =Dpr t Yt @Yt @Dpr t. Note that "DY t>0in the short run, because under nominal rigidities output is demand-determined. We cannot make unambiguous statements on "DY in the long run, for two fundamental reasons. First, we do not have any a-priori on the cyclical position of the economy; a de…cit-…nanced push in aggregate demand might have long-term e¤ect if used to close output gap. Second, as mentioned in the Introduction, we disregard the distinction between expenditure and revenue components of the primary de…cit; while the former - assuming non-productive public spending - has no real e¤ects in the long run, the latter might do. It is crucial to remind that "DY tdoes not coincide with the standard de…nition of the …scal policy multiplier, which normally measures the output level e¤ect resulting from level changes in …scal variables2@Yt @Dpr t. Empirical estimations of @Yt @Dpr toscillates around unity3for government spending multipliers and 1For the moment we assume a primary de…cit, so that it enters equation (1) with a positive sign. It would be more correct to reason in terms of primary budget position, implying the use of absolute value (displaying a de…cit or a surplus according to, respectively, the positive or negative values of Dpr). We specify that later on. 2Or, equivalently, the percentage change in GDP following a one per cent increase in the output share of the …scal variable. 3In recent literature we can …nd estimations around one (Barro and Redlick 2010, Guajardo 3 vary considerably when it comes to tax multiplier4. Moreover, recent research (Auerbach and Gorodnichenko 2012, Corsetti et al 2010, Batini et al 2010) has emphasized that …scal policy multipliers vary according to the cyclical position of the economy. The elasticity "DY tbeing based on percentage change as some of the empirical studies - is even more dependant on the state of the economy, as it is the product of the multiplier and the primary balance-to-GDP ratio. As the latter is often very modest, it is very unlikely that, regardless the multiplier’s size, the elasticity is greater than one. After a simple manipulation equation (2) becomes: @Bt Yt @Dpr t =#tBt1 Dpr tYt"D# "DY t1 Yt"DY t1(3) where we denote by "D# =Dpr t #t @#t @Dpr tthe elasticity of the average cost of debt to primary budget position. Regarding the sign of "D#, no unambiguous theoretical a priori can be made. However many empirical contributions (Bernoth et al 2004, Ardagna et al 2007, Laubach 2009) point out a positive relationship between deterioration of budget position and the average cost of debt, through the increase in risk premia occurring as a result of the increase in credit risk. Let’s now study the sign, assuming that the budget displays a primary de…cit (Dpr t>0) : If (3) is positive, it means that a de…cit reduction succeeds in causing a reduction in debt-to-GDP ratio. We call this scenario "pro-austerity", in order to indicate that a given decrease in primary budget is e¤ective in reducing the stock of government liabilities as a ratio to nominal income. If, on the other hand, (3) is negative, then a de…cit reduction increases the debt-to-GDP ratio and the budget consolidation results to be ine¤ective. We label this scenario "anti-austerity". pro austerity : @Bt Yt @Dpr t >0(4) anti austerity : @Bt Yt @Dpr t <0(5) Obviously when (3) is equal to zero, then changes in primary de…cit do not a¤ect debt-to-GDP ratio. In that case the three e¤ects recalled in the introduction o¤set each other. et al 2010, Ramey 2011 and Hall 2009) and some above one (Blanchard and Perotti 2002, Monacelli et al 2010, Blinder and Zandi 2010, Acconcia et al 2011, Fragetta and Melina 2011). 4Romer and Romer 2010, using the so-called "narrative approach", …nd it to be larger than three. 4 Looking at expression (3) it is easy to spot the su¢ cient condition for the prevailence of the pro or anti austerity regime. A budget consolidation is certainly successful if j"DY tj<min j"D# tj;1, whereas the "anti-austerity" regime dominates if j"DY tj>max j"D# tj;1. Given that, as noted above, it is quite unlikely that those elasticities are greater than one, those conditions basically tell us that a primary budget consolidation is certainly e¤ective if the elasticity of nominal output to primary de…cit is lower that the elasticity of the average cost of debt. Otherwise, the output loss associated to a budget consolidation is too large to allow for an actual reduction of the debt-to-GDP ratio. In order to compute the necessary and su¢ cient conditions for the prevailence of the pro/anti austerity regime it is necessary to study the sign of (3): @Bt Yt @Dpr t >0)#tBt1 YtDpr t"D# t"DY t1 Yt"DY t1>0 which means: Bt1 #t Dpr t"D# t"DY t>"DY t1 Multiplying both sides by Dpr tit becomes: Bt1#t"D# t> "DY t(Dpr t+Bt1#t)Dpr t "D# >Dpr t Bt1#t + 1"DY tDpr t Bt1#t (6) Note that the term Dpr t Bt1#t+ 1 is the inverse of the share of total primary de…cit devoted to interest payments on past debt. To account for the possibility that the primary budget position displays a surplus Spr t(= Dpr t)rather than a de…cit, the above conditions can be expressed in absolute values. We therefore state that a given change in primary budget is successful in creating the expected change in debt-to-GDP ratio if: j"D# tj>Dt Bt1#t j"DY tj  Dpr t Bt1#t (7) and it produces the opposite e¤ect if, instead: j"D# tj<Dt Bt1#t j"DY tj  Dpr t Bt1#t (8) In other words, the e¤ectiveness of a …scal policy aimed at reducing the stock of government liabilities relative to nominal income depends on the position of "D# with respect to a thresold "=Dpr t+Bt1#t Bt1#t"DY tDpr t Bt1#t Such a threshold can be drawn in a space "D# t"DY t(Figure 1): 5 A given combination of the two elasticities allows an actual debt-to-GDP reduction after a primary budget consolidation only if it falls in the "pro-austerity" space. The threshold line "represents the locus of elasticities "DY tand "D# t that leave B Yunchanged. The slope of "is given by Dpr #Bt1+ 1, from which we can see that the line is downward sloping if the government accumulates a primary surplus larger than interest payments (that is, if there is a budget surplus). Furthermore, Figure 1 clearly shows that even if the values of "DY tand "D# t are the same across two (or more) economies, the e¤ect of a one per cent change in primary budget can produce opposite e¤ects in terms of the debt-to-GDP ratio dynamics. Figure 2 replicates Figure 1 with data (year 2011) on the four major EMU economies and the overall Eurozone: We can observe that in 2011 Spain had the smallest "pro-austerity" space, whereas Germany had the widest. Hence a coordinated …scal policy move (even in presence of similar values of elasticities "DY tand "D# t) can have opposite e¤ects on their respective debt-to-GDP ratios if we are anywhere in the space between the two lines. 6 3 What can change the pro-austerity space? Figure 2 clearly shows that the threshold line "can considerably di¤er across similar countries, according to their budget position. This section analyzes the framework outlined in section 2, focusing on how di¤erent structure of budget de…cits at any point in time can increase or decrease the likelihood of a budget consolidation plan’s success. The partial derivative of the threshold "with respect to interest payments reads: @" @#tBt1 =Dpr #tBt1 (1 "DY t)(9) which shows us that countries with higher interest payments on the existing stock of debt see their pro-austerity space decrease for any "DY t<1and increase for any "DY t>1: In other words, if a country has an higher interest rate payment, a given reduction in primary budget de…cit has a lower (higher) chance to succeed if nominal output elasticity to primary de…cit is smaller (greater) than one. What is the intuition of that result? In order to understand it, let us recall the fundamental equation (2), written in a way that can best emphasize the three transmission channel from Dpr tto Bt Yt: 7