sharpe (2014) a modern money perspective on financial crowding out

A Modern Money Perspective onFinancial Crowding-out

TIMOTHY P. SHARPENewcastle Business School, Australia

ABSTRACT The withdrawal of discretionary fiscal stimulus and a renewed emphasis oninstitutional and ‘self-imposed’ budgetary constraints are evidence that the imperativeof fiscal sustainability and sound accounting fundamentals continue to drive fiscalpolicymaking within many advanced economies. To buttress the urgency for fiscalsustainability, neo-liberals often draw upon financial crowding-out theory. Despite anextensive literature, empirical applications are often misspecified due to their failure toaccount for different institutional arrangements. However, the policy responses ofnational governments to the Global Financial Crisis have highlighted the institutionaldisparities, presenting a unique opportunity for a rigorous empirical investigation. Thispaper develops panel vector error correction models for both sovereign and non-sovereign economies over the period 1999 to 2010 to examine financial crowding-out.The empirical evidence reveals crowding-out effects in non-sovereign economies, butnot within sovereign economies.

1. Introduction

As the nascent recovery following the Global Financial Crisis (GFC) remainstentative, policymakers within many advanced economies are pursuing fiscalconsolidation programmes as government debt and deficits reach ‘unsustainable’levels (see IMF, 2010a). A policy agenda geared to sustainable fiscal balancescharacterises the neo-liberal approach to the conduct of fiscal policy, which isclosely associated with the policy prescriptions of key Inter-GovernmentalOrganisations (IGOs), namely the International Monetary Fund (IMF) and theOrganisation for Economic Cooperation and Development (OECD) (seeSharpe & Watts, 2012).1 The withdrawal of discretionary fiscal stimulus and a

Review of Political Economy, 2013

Vol. 25, No. 4, 586–606, http://dx.doi.org/10.1080/09538259.2013.837325

Correspondence Address: Timothy P. Sharpe, Newcastle Business School, The University ofNewcastle, University Drive, Callaghan, NSW 2308, Australia. Email: [email protected] neo-liberal economic policy agenda, defined by Wacquant (2001, p. 404), consists of threepillars: ‘erasing the economic state, dismantling the social state, strengthening the penal state:these three transformations are intimately linked to one another and all three result essentiallyfrom the conversion of the ruling classes to neo-liberal ideology.’ The Washington Consensus pro-vides a useful summary of neo-liberal polices (see Williamson, 1990).

# 2013 Taylor & Francis

renewed emphasis on institutional and ‘self-imposed’ budgetary constraints areevidence that the imperative of so-called ‘fiscal sustainability’ and ‘sound’accounting fundamentals continue to drive fiscal policymaking within advancedeconomies.2

To buttress the urgency of fiscal sustainability, neo-liberals often assert thatrising budget deficits and public debt will cause higher interest rates (known asfinancial crowding-out) which can subsequently affect interest sensitive expendi-tures and hence economic growth. While these alleged effects of fiscal policy havebeen widely investigated within the theoretical and empirical literature, the policyresponses of national governments to the GFC have highlighted institutional dis-parities among advanced economies. These differences are particularly apparentwhen one defines a sovereign and non-sovereign economy, although the financialcrowding-out literature fails to make the distinction.

According to Modern Monetary Theory (MMT), a sovereign economy issuesits own fiat non-convertible currency and operates with a flexible exchange rate(Mitchell & Muysken, 2008). A non-sovereign economy does not possess eitherof these characteristics (e.g. Eurozone members). In responding to the GFC,many non-sovereign economies have been required to pursue harsh fiscal austeritymeasures, and have faced high interest rates on government bonds (e.g. Greece).However, sovereign economies (e.g. US, UK, Australia and Japan) have engagedin counter-cyclical fiscal measures despite the fact that, in some of these econom-ies, levels of public debt relative to GDP during the crisis have been similar to, orexceeded those of Eurozone members (e.g. Japan).

The impact of the GFC on national economies has demonstrated that sover-eign economies are endowed with greater freedoms in policy design andimplementation than non-sovereign economies, despite IGOs insisting that alladvanced economies must pursue medium-term fiscal consolidation strategies.Thus, institutional arrangements appear to have important implications for boththe conduct of fiscal policy and its impact on interest rates.

The objective of this paper is to investigate financial crowding-out,taking account of differences in institutional arrangements among advancedeconomies. Balanced panel data vector error correction models (VECMs) aredeveloped for both sovereign and non-sovereign economies over the period1999 to 2010. Furthermore, a number of deficit and public debt proxies areconsidered.

The remainder of this paper is organised as follows. Section 2 presentsfinancial crowding-out theory from the perspective of two competing theoreticaldiscourses; the neo-liberal and modern money perspectives. The extant empiricalliterature is also briefly reviewed. Section 3 discusses the data and model specifi-cation. Empirical results are reported and discussed in Section 4. Concludingremarks are presented in Section 5.

2Fiscal sustainability is often defined by reference to the algebra of debt and deficit dynamics (seeBlanchard et al. 1990; Buiter, 2010; Watts & Sharpe, 2013). Notwithstanding this, sound publicfinance and fiscal sustainability are not operational concepts.

A Modern Money Perspective on Financial Crowding-out 587

2. Financial Crowding-out

2.1. A Neo-liberal Perspective

Financial crowding-out theory has received wide coverage within the academicliterature and financial media, and is often referred to within the policy prescrip-tions of key IGOs.3 Most recently, the theory has been drawn on by neo-liberals topromote expansionary fiscal contractions and fortify the imperative of fiscal sus-tainability within many advanced economies.

In essence, the expansionary contractions hypothesis claims that expansion-ary outcomes, in terms of increased output growth, increased consumption andinvestment expenditure and lower unemployment rates, can be achieved via afiscal consolidation (reduction in government spending and/or an increase intaxes) over the medium to long term (see Alesina, 2010; Giavazzi & Pagano,1990). Financial crowding-out and ‘credibility’ effects are positioned as keydemand-side mechanisms, inducing private investment and economic growthunder conditions of fiscal withdrawal (Briotti, 2005). The OECD and IMF fre-quently draw on the expansionary contractions hypothesis to validate fiscal con-solidations within many advanced economies (see IMF, 2010b, 2010c; OECD,2009, ch. 3; 2011, ch. 4).4

In a report entitled Preparing Fiscal Consolidation, the OECD (2010, p. 7)presents three arguments justifying fiscal consolidations, all of which areembedded in the fiscal sustainability literature. First, ‘[h]igh public indebtednesscurtails the scope for countercyclical fiscal action, when needed’. Second,‘[b]allooning outlays on debt service could crowd out government spending ongrowth enhancing programmes’. Finally, ‘[e]xcess supply of government bondsmay put upward pressure on interest rates and therefore crowd out private invest-ment and compromise long-term output’. The first and second propositions haveties with the emerging literature on fiscal space (see Heller, 2005). The thirdjustification for fiscal consolidation relates to ‘textbook’ financial crowding-outtheory. Culbertson (1968, p. 463) characterises the ‘textbook’ view, to whichthe OECD and IMF subscribe:

Fiscal policy provides additional spending in a world of sparse spending oppor-tunities. But it does not provide a new source of finance in a world where spend-ing is constrained by sources of funds of finance. The government expendituresare financed in debt markets in competition with private expenditures. The caseleast favorable to fiscal policy is that in which the additional government

3Crowding-out arguments opposing government intervention can be traced back to Adam Smith,John Stuart Mill and J.B. Say. The synthesis of Keynes’s General Theory into the IS/LM frameworkby John R. Hicks and Alvin Hansen created more empirical interest in the impact of governmentspending on interest rates, and subsequently on investment and income. For an overview, seeSpencer & Yohe (1970) or Bernheim (1989).4Amid weak GDP growth, IMF (2012a) has now largely abandoned the expansionary fiscal contrac-tions argument. Further, IMF (2012a, p. 41) concedes that ‘the [fiscal] multipliers used in generatinggrowth forecasts have been systematically too low since the start of the Great Recession, by 0.4 to1.2. . .. [A]ctual [fiscal] multipliers may be higher, in the range of 0.9 to 1.7’.

588 T.P. Sharpe

borrowing simply crowds out of the market an equal (or conceivably evengreater) volume of borrowing that would have financed private expenditures.

Thus, increased net government spending requires increased government debtto finance the additional outlays. Then, on the presumption that the supply ofcredit is constrained, government debt must compete with private debt, ineffect, putting upward pressure on market interest rates. Additionally, increasedgovernment spending reduces overall national savings (defined as both privateand public sector savings) and therefore the supply of loanable funds, whichalso results in higher interest rates. Higher interest rates are expected to increaseborrowing and debt servicing costs, thus crowding-out private investment spend-ing. Furthermore, ‘ever larger deficits,’ so the argument goes, ‘generate everhigher interest rates, as the government must offer incentive to encourageprivate lenders to accept its IOUs in exchange for their saving or as a premiumagainst the risk of default or—again, worse still—the possibility of future mone-tisation to “repay” the deficits’ (Fullwiler, 2006, p. 4).

A consequence of the GFC has been increased public debt and deficit levelsamong many advanced economies, largely due to the workings of automaticstabilisers with private sector demand falling and unemployment rising. Policy-makers fear that these levels of public debt and deficits are unsustainable, andso, must be reduced. Thus, a policy consensus, reinforced by the prescriptionsof IGOs has emerged that requires the medium-term pursuit of fiscal consolidationto bring public debt levels back to sustainable levels and reduce interest rates (seeSharpe & Watts, 2012). Notwithstanding this, there seem to be some disparitiesemerging within the economic data that financial crowding-out theory cannotreconcile. This threatens the legitimacy of the argument as a justification forfiscal consolidation among many advanced economies. These disparities are illus-trated by Figures 1 to 4.5

Figures 1 and 2 illustrate the impact of the GFC since 200722008 on thegross financial liabilities (debt) to GDP ratios in a number of advanced economies.Some countries (Ireland, for example) have experienced more dramatic increasesin gross debt than others, which was largely the result of government interventionto recapitalise the banking system. Notwithstanding this, increased governmentdeficit and debt ratios can be expected in the presence of declining privatesector spending. Differences among these advanced economies are illustratedby the behaviour of long-term government bond rates in response to the GFC.Government bond yields within Eurozone economies (Figure 3) have generallyincreased with gross government debt. On the other hand, government bondyields have remained reasonably stable, or have even declined, within non-Euro-zone economies (Figure 4) over the same period. At face value, financial crowd-ing-out theory cannot reconcile these outcomes.

Proponents of expansionary contractions may draw on the ‘credibility’ argu-ment in an attempt to rationalise the above phenomena. That is, ‘[i]f agents believe

5The data for all figures are drawn from OECD Economic Outlook No. 90, Annex tables, availableat: http://www.oecd.org/eco/outlook/economicoutlookannextables.htm.


that the [fiscal] stabilization is credible and avoids a default on government debt,they can ask for lower premiums on government bonds’ (Alesina, 2010, p. 4). Fur-thermore, ‘[i]f a country does not have fiscal credibility, it will have much moredifficulty in continuing its fiscal programme because creditors will refuse to con-tinue lending or will demand a higher risk premium’ (OECD, 2010, p. 5).

Figure 1. Eurozone economies: gross financial liabilities to GDP.

Figure 2. Non-Eurozone economies: gross financial liabilities to GDP.

590 T.P. Sharpe

However, this argument is inconsistent with the data and the economic commen-tary among sovereign non-Eurozone economies.

Eurozone economies have pursued relatively more severe fiscalausterity measures than non-Eurozone economies. In most instances, these

Figure 3. Eurozone economies: long-term government bond yields.

Figure 4. Non-Eurozone economies: long-term government bond yields.


programmes have identified clear medium-term frameworks and, in some cases,binding short- and longer-term targets (i.e. conditions of IMF/ECB/EC supportedloans). Yet, these economies continue to face rising long-term government bondrates.

On the other hand, the non-Eurozone economies depicted in Figures 2 and4—Canada, Japan, the UK and the US—have pursued limited attempts at conso-lidation (see OECD, 2011, ch. 4). Furthermore, the ‘fiscal credibility’ of the USshould have been undermined by the recent political debate over the ‘debtceiling’ (a politically imposed constraint on government spending), allegationsof a default, and the historical credit rating downgrade on US government debtby Standard & Poor; yet long-term interest rates for the US remain at recordlows (see also the UK). Japan has experienced decades of allegedly profligate gov-ernment spending, multiple downgrades on public debt by major credit ratingagencies and currently has the highest gross debt ratio in the world (now over200% of GDP) but continues to attract long-term rates on government bondsbelow 2%.6

The disparities among government bond yields of advanced economiesostensibly challenge the financial crowding-out and ‘credibility’ arguments ofexpansionary contraction theorists. Until recently, fundamental institutionaldifferences between Eurozone and non-Eurozone economies have been largelyunnoticed, overshadowed by the noisy neo-liberal rhetoric in favour of ‘sound’public finances (see Watts & Sharpe, 2013). In contrast to financial crowding-out theory, modern monetary theory makes the necessary distinction between asovereign and a non-sovereign economy, and in doing so, offers an explanationconsistent with the stylised evidence presented above.

2.2. A Modern Money Perspective

Modern Monetary Theory (MMT) offers an ‘accounting-consistent, operationally-sound theoretical approach to understanding the way fiat monetary systems workand how policy changes are likely to play out’ (Mitchell, 2011). The conduct offiscal policy according to MMT counters the dominant neo-liberal perspectivewhich is driven by accounting imperatives rather than advancing publicpurpose.7 The principles of MMT make it possible to distinguish between a sover-eign and non-sovereign economy. The former issues a non-convertible fiat cur-rency and operates with a flexible exchange rate, so that monetary policy isfreed from the need to defend foreign exchange reserves and maintain a desiredexchange parity, examples being Australia, the US, the UK, Japan and Canada(Mitchell, 2009; Mitchell & Muysken, 2008). A non-sovereign economy does

6This is often explained by noting that most Japanese government debt is held domestically. Butwhat matters for borrowing rates is the currency in which the debt is denominated, not the nationalityof the debt-holders (see the next section).7Mitchell (2009, p. 11) outlines the broad charter for advancing public purpose, which includes, ‘fullemployment and price stability, poverty alleviation and environmental sustainability. . .’ Here, fullemployment is defined as 2 percent unemployment, no hidden unemployment, and no underemploy-ment (see also Mitchell & Muysken, 2008).

592 T.P. Sharpe

not possess either of these characteristics; examples are France, Spain, Portugal,Ireland and Greece.

Financial crowding-out theory was initially proposed and analysed in thecontext of a convertible currency system, that is, the gold standard and theBretton Woods fixed exchange rate agreement (1946271). Mitchell (2009,p. 9) notes that ‘[e]conomic policy ideas that prevailed in the previous monetarysystems (based on convertibility) are no longer applicable in a fiat monetarysystem’.

MMT principles can be applied to inform Culbertson’s (1968, p. 463) ‘text-book’ view of financial crowding-out. First, ‘spending is constrained by sources offunds of finance’. For the non-government sector, credit is constrained by the per-ceived creditworthiness of the borrower. A bank will lend to anyone it discerns asable and willing to fulfil the obligations of the loan, although underwriting stan-dards may vary over the business cycle. Government spending within sovereigneconomies is not constrained by sources of finance. In a fiat currency system‘money is a creature of the state’ (Lerner, 1947, p. 313). While it has no intrinsicvalue it is accepted as a means of exchange because the government requires ‘itsuse to relinquish private tax obligations to the state’ (Mitchell, 2009, p. 9). Then, ifconvertibility is abandoned and the government is the monopoly issuer of the fiatcurrency, government spending is no longer revenue-constrained and government‘saving’ is impossible. In a sovereign economy, government spending is only con-strained by the availability of real goods and services, denominated in the nationalcurrency (Mitchell, 2009; Mitchell & Muysken, 2008). On the other hand, govern-ment within non-sovereign economies are revenue-constrained since they are notthe monopoly issuer of the currency accepted in payment of taxes. Consequently,government spending must be financed via taxation or borrowing in debt markets.

Second, ‘government expenditures are financed in debt markets in compe-tition with private expenditures’. Recall the first point, government spendingwithin sovereign economies does not require ‘financing’. Instead, governmentdebt serves at least two distinct purposes.8 Government debt is primarily usedas an interest-rate maintenance mechanism within the interbank market (see Full-wiler, 2006; Mitchell, 2009; Mitchell & Muysken, 2008). Ceteris paribus, deficitgovernment spending increases the reserves within the interbank market viaincreased deposits, which places downward pressure on the interbank lendingrate. This can potentially undermine the target policy rate set by the centralbank. Thus, in the presence of a ‘corridor’ system and a desire to maintain a posi-tive policy rate, the central bank must drain these ‘excess reserves’ by selling,typically under a repurchase agreement, an interest-bearing asset (governmentdebt); otherwise the interbank lending rate will fall towards the support rate, orzero if no support rate is offered.9 If all bank reserves are remunerated at the

8Government debt may also be desired by financial markets because it represents a risk-free interest-bearing asset, which can be used as a benchmark for pricing other debt instruments or financial secu-rities, and to balance the risk structure of investment portfolios; see Mitchell (2009) for a critique.9This point is often explicit in central bank documentation, in particular the Reserve Bank ofAustralia’s account of the conduct of Open Market Operations (see RBA, 2011).


interbank target rate (i.e. the target rate equals the support rate), or the central bankhas a desire to let the interbank rate fall to zero, government bonds do not need tobe issued for this purpose. Hence, contrary to the financial crowding-out argu-ment, government deficits can reduce short-term interest rates, although only ifdesired by the government and the central bank.

Government debt also serves a political purpose within sovereign economies.Many advanced sovereign economies have imposed voluntary constraints on gov-ernment spending which resemble the formal, although voluntarily adopted, con-straints of non-sovereign Eurozone economies. These arrangements largely serveto promote fiscal discipline by matching government spending dollar-for-dollarwith borrowings from the non-government sector (Mitchell, 2009). These volun-tarily adopted guidelines, albeit with varying degrees of legal and political clout(Gutierrez & Revilla, 2010), have fostered the neo-liberal mantra, whicherroneously likens the financial constraints of government to those of a household.In this vein, governments that reduce their outstanding stock of debt and targetbudget surpluses are viewed as prudent managers of taxpayers’ funds. Thisview of the role of government has now become so deeply embedded in politicaldiscourse that attempts to offer alternative policy are quick to be dismissed by thegeneral public.

The above discussion has outlined two competing discourses of financialcrowding-out. The recent behaviour of long-term bond rates within sovereigneconomies, despite allegedly ‘unsustainable’ public debt levels, has potentiallyexposed the importance of institutional arrangements for the conduct of fiscalpolicy. From an econometric perspective, the disparity in long-term bond ratessuggests, at the very least, that it is important to distinguish between sovereignand non-sovereign economies.

2.3. Empirical Evidence

Financial crowding-out is allegedly transmitted from increased governmentdeficit/debt ratios to higher interest rates, and subsequently, reduced interest-sensitive expenditures. These linkages are often treated separately within theempirical literature. The literature outlined below, investigates the first linkage,which is the focus of this study. Clearly if interest rates are not responsive toproxies of fiscal policy (deficits or debt), the second linkage becomes irrelevant.In any case, the interest sensitivity of private investment is widely disputed (seeHemming et al., 2002).

The financial crowding-out empirical literature is mixed and highly conten-tious. A survey of 59 papers by Gale & Orszag (2003) reports that 29 found sig-nificant effects of deficits on interest rates, 19 found insignificant effects and 11returned with mixed results. Surveys of the financial crowding-out literature byEuropean Commission (2004), Gale & Orszag (2003) and OECD (2009) report,on average, a 1% of GDP deterioration in the fiscal balance raises interest ratesby 30–60 or 15–80 basis points, and long-term interest rates in particular by10–60 basis points. OECD (2009) reports that an increase in the stock of debtby 1% of GDP raises long-term interest rates by 3–50 basis points. IMF(2010b) reports that output is expected to increase by 1.4% and real interest

594 T.P. Sharpe

rates are expected to decline by 30 basis points, in the long-term, in response to areduction of 10 percentage points in the debt-to-GDP ratio in Japan, the Euro areaand the US.

Panel data investigations however often fail to separate sovereign and non-sovereign economies and are consequently misspecified. While studies comparingtime-series results of individual sovereign and non-sovereign economies, particu-larly recent investigations, reveal some disparity in regression results, no referenceis made to differences in underlying institutional arrangements. Rather, a findingthat rising government debt reduces long-term interest rates is often reconciled byreference to a portfolio/liquidity effect (Caporale & Williams, 2002; Marattinet al., 2011).

3. Data Analysis and Model Specification

3.1. Data

To investigate financial crowding-out, equation (1) is estimated for two balancedpanels, one containing sovereign economies, and the other with non-sovereigneconomies:

yit = a0 + a1Zit + a2Xit + 1it i = 1, 2, ...N; t = 1, 2...T (1)

In equation (1), yitis the long-short interest rate spread (SPR). Long-term interestrates are 10-year government bond rates, while short-term rates are either thecentral bank policy rate or interbank cash/call rate. Long-term interest rates aregenerally considered more relevant to investment decisions, and are lessexposed to the monetary policy actions of central banks. The spread betweenlong- and short-term interest rates is used here because it effectively accountsfor short-term interest rate or monetary policy effects (see Ardagna et al., 2004;Paesani et al., 2006). Zit is a vector of fiscal proxies, including: general govern-ment cyclically-adjusted balance to potential GDP (CAFB) (deficit is positive);10

expected deficit (see Idier et al., 2007; OECD, 2009), derived as the average ofthe general government (interest inclusive) financial balance to GDP over thenext 5-years (EDEFICIT) (deficit is positive); general government gross financialliabilities to nominal GDP (GDEBT); and a measure of debt service (see OECD,2009), derived as a ratio of general government net debt interest payments togeneral government total tax and non-tax receipts (DSER). Squared gross debt(GDEBTSQ) and debt service (DSERSQ) terms are also included to account forpossible non-linear effects (see Ardagna et al., 2004; Bernoth et al., 2004). Xit

is a vector of other explanatory variables including: the inflation rate (P), thereal-effective CPI based exchange rate (REER), log of the share price index(SPI), and current account balance to GDP ratio (CAB) (deficit is negative). 1it

is an error term.

10See OECD (2006) for the methodology. Since the revision of the Stability and Growth Pact in2005, cyclically-adjusted measures are given more weight than headline measures in the EUfiscal surveillance framework (Larch & Turrini, 2009).


For the sovereign panel, we used data from Australia, US, UK, Japan andCanada. Non-sovereign economies include France, Spain, Portugal, Ireland andGreece.11 Given the short time dimension, 1999 to 2010, panel data analysisallows those data to be used more efficiently than time-series estimations. Theintent of the empirical work is to identify whether the effect of selected fiscal vari-ables on long-term interest rates are statistically significant within sovereign andnon-sovereign economies. The selected variables reflect the extant empirical lit-erature and data availability, with a particular emphasis on recent developmentsconcerning fiscal proxies.12

Sovereign models also include a credit risk spread (CR), derived as the differ-ence between the 12 month average of daily selected long-term corporate bondrates and 10-year government bond rates (see OECD, 2009). Non-sovereignmodels include the German benchmark 10-year bond rate (BB).13 Germanbonds are generally considered high-quality and low-risk. Ceteris paribus, anincrease in demand for low-yield German bonds may signal a loss of confidencein the wider Eurozone bond market (see De Grauwe, 2011). Consequently,German long-term rates are kept significantly lower than those of other Eurozonemembers. A priori, an inverse relationship is expected between the German bench-mark bond rate and the long-short spread of other Eurozone economies if a liquid-ity/portfolio effect is present.

3.2. Panel Unit Root Tests

Prior to estimating possible long-run cointegrating relationships among the vari-ables, the order of integration must be established. The short time dimension ofthe data set can distort the power of individual unit root procedures (Christopoulos& Tsionas, 2004), thus LLC, IPS and ADF-Fisher panel unit root tests are com-monly applied. These tests are based on the conventional ADF unit root specifica-tion (2).

Dyit = a0i + riyit−1 +∑ pi

j=1wijDyit−j + 1it i = 1, 2, ...N; t = 1, 2, ...T (2)

The null hypothesis is a unit root process (H0: ri = 0) against the alternative of astationary process (H1:ri , 0) for each individual series. The LLC test,however, is problematic as it assumes homogeneous autoregressive coefficients,i.e. ri = r for all i, which is highly restrictive. The Maddala & Wu (1999)

11Recent trends in the data suggest that the implications of MMT are likely to be more pronouncedfor this particular subset of non-sovereign economies. The regressions commence in 1999 tocoincide with the introduction of the euro.12Data have been obtained from the OECD Economic Outlook Annex tables, IMF InternationalFinancial Statistics and Fiscal Monitor (IMF, 2011), Datastream, and relevant Central Bank data-bases. All regressions are performed in Eviews 7.13Appropriate corporate bond yields were unavailable for selected non-sovereign economies.Default proxies are only applicable to non-sovereign economies. Sturzenegger (2002, p. 18) cautionsagainst the use of credit default swaps, suggesting that markets can, at times, become ‘extremelythin’ which can create a ‘distorted view of default probabilities’. Notwithstanding this, data arelimited.

596 T.P. Sharpe

Fisher-type test aggregates the individual unit root p-values from each cross-sec-tional unit to test for a unit root in the combined series. Following Maddala & Wu,if the test statistics are continuous, the significance levels pi (i ¼1, 2, ... N ) areindependent uniform (0, 1) variables, and −2 logpi has a x2

(2) distribution. Thecombined p-value, defined by equation (3), is obtained from the additive propertyof x2 variables which has a x2 distribution with 2N degrees of freedom.

pl = −2∑N

i=1logpi (3)

Maddala & Wu (1999) use simulations, imposing various assumptions of station-ary processes to find that the Fisher test is superior to LLC and IPS given its highpower and low size distortions. However, the LLC, IPS and Fisher tests assumecross-sectional independence which is likely to be violated in this study. There-fore, the test statistics are no longer valid. Nevertheless, these tests are stillapplied within the empirical financial crowding-out literature. To overcomethis, Maddala & Wu propose a bootstrap method, based on the Fisher test, toderive the empirical distributions of the test statistics. In essence, a bootstrapsample is generated under the null of a unit root (ri = 0) by drawing with replace-ment from the residual series obtained from regression equation (2). However,1it

cannot be re-sampled directly since the errors are potentially cross-correlated, thus1it = [11,t, 12,t, ..., 1N,t] are re-sampled to retain the cross-correlation structure ofthe error term. This process is repeated 2,000 times to obtain the MW-Fisher teststatistic.14

For sovereign economies, the MW Fisher-Bootstrap results suggest that allvariables are stationary in levels, i.e. I(0), except GDEBT and GDEBTSQwhich are I(1) or stationary in first differences. For the non-sovereign economies,BB, GDEBTSQ, DSER and DSERSQ are I(1), all other variables are I(0).15

3.3. Panel Cointegration Tests

The results of the panel unit root tests reveal a mixture of I(0) and I(1) variables,thus conventional panel cointegration tests are not appropriate. This study uses abounds test approach to determine possible level relationships among the variablesthat can be applied irrespective of whether the regressors are purely I(0), purelyI(1) or a mixture of the two. Pesaran et al. (2001) have developed asymptotic criti-cal value bounds for the F-statistic to test the joint significance of the lagged levelsof the variables. If the computed F-statistic exceeds the upper critical bound, thenull hypothesis of no level (cointegrating) relationship among the variables isrejected. If the F-statistic is below the lower critical bound, we fail to reject thenull in favour of no cointegration. The result is inconclusive if the computedF-statistic falls within the critical value bounds.

Panel cointegration tests are often favourable since they ‘combine the abilityof time series studies to yield causality inferences with the increase in sample size

14Eviews program code for these calculations was obtained from Galimberti & Cupertino (2009).15Panel unit root results are available from the author upon request.


afforded by using cross-sectional data’ (Christopoulos & Tsionas, 2004, p. 71). Anadvantage of the bounds test with panel data is the relative ease of accounting forcross-section and period effects. Failure to address these issues would result inmisspecification. Thus, the bounds cointegration specification of equation (1) isestimated with cross-section and/or period dummy variables to capture theseeffects.16 The bounds specification (with no fixed effects) of equation (1) isexpressed as follows.

Dyit = a0 + a1yit−1 + a2Zit−1 + a3Xit−1

+∑k

j=1a4Dyit−j +

∑n

m=0a5DZit−m +

∑q

p=0a6DXit−p + 1it (4)

Optimal lag lengths on the differenced regressors are based on the Akaike infor-mation criterion (AIC) (see Pesaran et al., 2001). Each variable is specified as thedependent variable, and the F-statistic is computed to test the null hypothesis of nocointegration (5).

H0:a1 = a2 = a3 = 0 (5)

Panel cointegration results (with SPR as the dependent variable) confirm long-runcointegrating relationships among all fiscal proxies for both sovereign and non-sovereign economies (with and without fixed effects).17

3.4. Model Specification

Cointegrating relationships among the variables in equation (1) have been ident-ified. A panel vector-error correction model (VECM) can now be estimated thatalso allows for potential endogenous explanatory variables. Unlike other econo-metric methods such as 2SLS, 3SLS and GMM, a VECM does not ignore the inte-gration and cointegration properties of the underlying data, which can be used tomake inferences on the long-term relationships among the variables. Christopou-los & Tsionas (2004, p. 60) add, ‘it is not clear that the estimated panel models [i.e.GMM] represent a structural long run equilibrium relationship instead of a spur-ious one’. A VECM is a restricted VAR, which constrains the long-run behaviourof the endogenous variables to converge to their cointegrating relationships whileallowing for short-run adjustment dynamics. The VECM specification of equation(1) is as follows:

Dyit = a0 +∑k

j=1a1Dyit−j

+∑n

m=0a2DZit−m +

∑q

p=0a3DXit−p + a4ecmt−1 + vit (6)

All variables and vectors are as previously defined. The error correction term(ecmt21) corrects deviations from the long-run equilibrium through a series of

16Pesaran et al. (2001) asymptotic critical values may be affected by the inclusion of cross-sectiondummies. Period dummies are for the GFC period 2008–2010.17Panel cointegration results are available from the author upon request.

598 T.P. Sharpe

short-run partial adjustments. vitis an error term. The optimal lag length on the dif-ferenced regressors is based on the AIC (see above).

However, it is well known that least squares dummy variable (LSDV) esti-mations yield biased and inconsistent results when applied to a dynamic paneldata setting since, by construction, the lagged dependent is correlated with theerror term (see Judson & Owen, 1999; Nickell, 1981). To overcome this, Anderson& Hsiao (1982) recommend an instrumental variable (IV) estimation techniqueusing higher-order lags on the differenced dependent variable as instruments (seealso Christopoulos & Tsionas 2004). This IV technique is applied and diagnostictests are performed to ensure the models are well-specified, the errors are normallydistributed and serially uncorrelated, and the over-identifying restrictions are valid.For brevity, the most robust models according to these diagnostic tests are reported.

4. Empirical Results and Discussion

The regression results are presented in Tables 1 to 3. The error correction terms(ECM) are all significantly different from zero (at 1%) indicating the cointegrationresults are reliable. Diagnostic tests affirm that the models are data congruent andwithout specification error. Although well specified, the long-short spread insovereign economies may require different explanatory variables than non-sover-eign economies as indicated by the relatively low F-statistics. By failing to dis-tinguish between sovereign and non-sovereign economies, the extant literaturemay have overlooked appropriate macroeconomic variables.

The results have revealed robust differences among fiscal variables foradvanced sovereign and non-sovereign economies, which affirm the importanceof institutional arrangements. The IMF now concedes that ‘fiscal indicatorssuch as deficit and debt levels appear to be weakly related to government bondyields for advanced economies with monetary independence’ (IMF, 2012b,p. 38; emphasis added).

The cyclically-adjusted financial balance (CAFB), gross government debtratio (GDEBT) and the debt service ratio (DSER) have a positive and statisticallysignificant impact on the long-short interest rate spread (SPR) for non-sovereigneconomies but are insignificant for sovereign economies.18 The estimated elasti-cities for non-sovereign economies are 0.05, 0.02 and 0.29 respectively.GDEBT coefficients are similar in magnitude to those of Ardagna et al. (2004).In addition, non-linear effects of DSER and GDEBT are significant for non-sover-eign economies (see Ardagna et al., 2004; Bernoth et al., 2004). However, esti-mating the potential threshold effects of these non-linear variables is a directionfor further study (see Sharpe, 2013).

As suggested by Bernoth et al. (2004), the debt service ratio seems to capturemore of the variation in interest rate spreads among EU countries than the fiscal(deficit and debt) variables. Bernoth et al. (2004) also found that the debt

18Fiscal proxies are not statistically significant for sovereign economies. EDEFICIT is not significantfor non-sovereign economies. The results are similar where nominal government bond rates are usedas the dependent variable.


service indicator became a more important determinant of interest rates after theformation of the European Monetary Union (EMU). MMT argues that by joiningthe EMU a country relinquishes its fiscal-monetary sovereignty, and so canbecome insolvent. Hence, rising debt servicing costs may signal an increasedrisk of insolvency, which materialises as rising risk premiums on long-termdebt, potentially creating a positive feedback mechanism.19

Table 1. Deficit VECM Results

Sovereign economies Non-sovereign economies

VariableCoefficient(t-statistic)

Coefficient(t-statistic) Variable

Coefficient(t-statistic)


SPR(-1) 0.320385(0.697283)

20.010850(20.075113)

SPR(-1) 20.035055(20.256901)

0.871491**(2.265438)

CR 0.511455*(1.847552)

0.508343**(2.557915)

BB 20.433839(21.572457)

0.361071(0.813095)

CAFB 0.156816(0.863585)

BB(-1) 21.281997***(26.573157)

EDEFICIT 20.293342(21.758614)

CAFB 0.054687**(2.239606)

P 0.016443(0.144906)

20.143607(21.657589)

EDEFICIT 20.008431(20.096663)

REER 20.036536(21.571308)

20.048250***(22.745763)

P 20.330096***(24.792311)

20.572826***(25.102467)

SPI 0.304505(0.359055)

0.415984(0.746226)

REER 20.004038(20.149124)

0.118731(1.283858)

CAB 20.148337(20.845489)

20.206026*(21.922980)

SPI 21.841284***(24.176228)

22.648656***(23.223896)

ecm(-1) 20.581112***(23.288259)

20.577679***(24.210935)

CAB 20.050412(20.874663)

20.001891(20.021073)

ecm(-1) 20.477415***(23.773267)

20.781945***(23.666519)

Diagnosticsa Diagnostics�R

20.447827 0.620564 �R

20.943513 0.918016

Normalityx2

(2)

1.726720 4.468299 Normalityx2

(2)

0.192079 0.566378

AR(2) x2(2) 0.260508 3.878691 AR(2) x2

(2) 1.370934 0.092082F-statistic 3.388938*** 5.076173*** F-statistic 47.14221*** 28.55209***Specification 0.772008 0.552645 Specification 0.532711 1.387300SS x2

(q) 0.197733 0.060794 SS x2(q) 0.749329 n/ab

***,**,* significant at the 1%, 5% and 7% level respectively. All variables are in firstdifference. No fixed effects included. Instruments are lags of DSPR.aNormality: Jarque-Bera normality test. AR(2): Test for the null of no second-order serialcorrelation. F-statistic: Joint significance of coefficients. Specification: F-test for the nullof jointly irrelevant fixed effects. SS: Sargan statistic for the null of valid over-identifyingrestrictions (q).bJust identified.

19Buiter (2010) makes a similar claim, though he does not adequately distinguish between a sover-eign and non-sovereign economy.

600 T.P. Sharpe

Unlike the mainstream literature, an insignificant deficit or debt variable isnot interpreted as evidence in favour of Ricardian equivalence. That is, we donot conclude that the private sector is acting rationally to offset increased govern-ment spending by increasing current private savings in anticipation of higherfuture taxes.20 Rather, increasing fiscal deficits or debt within sovereign econom-ies does not raise liquidity or insolvency concerns, which tend to manifest

Table 2. Gross Debt VECM Results






SPR(-1) 0.391367(1.288793)

20.011469(20.031246)

SPR(-1) 0.028038(0.272266)

0.071135(0.656737)

CR 0.775135**(2.416351)

0.428627**(2.222836)

BB 20.498575**(22.115106)

20.077215(20.318792)

GDEBT 0.043272(1.043154)

0.109135(1.130739)

BB(-1) 21.153920***(27.149814)

20.965106***(25.491885)

GDEBTSQ 20.000267(20.909255)

GDEBT 0.022507**(2.380609)

20.027706(21.213027)

P 20.164651(21.053452)

20.092919(21.269077)

GDEBTSQ 0.000330**(2.170639)

REER 20.056198*(22.018890)

20.014449(20.532871)

P 20.374321***(26.300810)

20.304101***(24.243995)

SPI 0.548516(0.798528)

20.124751(20.285935)

REER 0.019303(0.775162)

20.007682(20.222837)

CAB 20.120408(20.833908)

20.035580(20.265352)

SPI 21.428634***(23.627605)

22.017442***(24.581051)

ecm(-1) 20.772144***(23.537225)

20.457135***(22.984946)

CAB 20.017178(20.353294)

0.057002(1.003809)

ecm (21) 20.497413***(24.329741)

20.427859***(22.909993)


20.582559 0.595858 �R

20.966288 0.940458

Normalityx2

(2)

3.461375 0.671304 Normalityx2

(2)

0.311372 0.778511

AR(2) x2(2) 0.571173 0.545326 AR(2) x2


(q) 1.267687 0.799585 SS x2(q) 0.514269 1.521704

***,**,* significant at the 1%, 5% and 7% level respectively. All variables are in firstdifference. No fixed effects included. Instruments are lags of △SPR.aNormality: Jarque-Bera normality test. AR(2): Test for the null of no second-order serialcorrelation. F-statistic: Joint significance of coefficients. Specification: F-test for the nullof jointly irrelevant fixed effects. SS: Sargan statistic for the null of valid over-identifyingrestrictions (q).

20Ricardian equivalence can be dismissed on the basis of its theoretical assumptions (see Bernheim,1989).


themselves as higher long-term interest rates. This is because financial marketparticipants understand the fundamental institutional differences, which arehighlighted in Modern Monetary Theory. Fund manager and investment strategist,Marshall Auerback (2010) notices ‘[t]he market’s recent response to the intensify-ing pressures in the euro zone suggests that investors are beginning to differentiatebetween countries . . . Whether consciously or not, the markets are demonstratingthat they understand the distinctions between users of currencies (who face anexternal funding constraint), and those nations that face no constraint in theirdeficit spending activities because they are creators of currency’ (see also DeGrauwe, 2011).

Table 3. Debt Service VECM Results






SPR(-1) 0.371701(0.984711)

0.221195(1.476541)

SPR(-1) 0.197074(1.464493)

0.236723(1.788941)

CR 0.591663**(2.160374)

0.538004**(2.249931)

BB 20.240386(21.008456)

20.180177(20.784047)

DSER 20.009018(20.033289)

0.435139(1.111088)

BB(-1) 21.080795***(26.201986)

20.964326***(25.587090)

DSERSQ 20.030057(20.520968)

DSER 0.295174**(2.684237)

0.060063(0.393578)

P 20.041736(20.373251)

20.149370(21.802663)

DSERSQ 0.012843**(2.202309)

REER 20.037389(21.581545)

20.037173**(22.410456)

P 20.368577***(25.768799)

20.396428***(26.611405)

SPI 20.144908(20.196534)

0.350221(0.894346)

REER 0.014696(0.536797)

20.003199(20.136452)

CAB 20.158950(20.858649)

20.175639(21.517867)

SPI 21.805440∗∗∗

(24.642780)21.769253***(25.317290)

ecm(-1) 20.704690***(22.726045)

20.639626***(23.186298)

CAB 0.002961(0.054437)

0.011711(0.239201)

ecm(-1) 20.574348***(25.441394)

20.550065***(25.562374)


20.447200 0.394208 �R

20.969506 0.978822

Normalityx2

(2)

2.218018 4.954195 Normalityx2

(2)

2.495697 2.159288

AR(2) x2(2) 0.331266 0.040235 AR(2) x2


(q) 0.439980 1.798843 SS x2(q) 0.602152 2.236528

***,**,* significant at the 1%, 5% and 7% level respectively. All variables are in firstdifference. No fixed effects included. Instruments are lags of DSPR.aNormality: Jarque-Bera normality test. AR(2): Test for the null of no second-order serialcorrelation. F-statistic: Joint significance of coefficients. Specification: F-test for the nullof jointly irrelevant fixed effects. SS: Sargan statistic for the null of valid over-identifyingrestrictions (q).

602 T.P. Sharpe

For ‘creators of currency’ such as the US, former Federal Reserve ChairmanAlan Greenspan (1997) has unequivocally acknowledged that ‘[c]entral banks canissue currency, a noninterest-bearing claim on the government, effectively withoutlimit. . . . That all of these claims on government are readily accepted reflects thefact that a government cannot become insolvent with respect to obligations in itsown currency. A fiat money system, like the ones we have today, can produce suchclaims without limit.’ Martin Feldstein has reiterated the point in a recent inter-view:

French officials apparently don’t recognise the importance of the fact thatBritain is outside the eurozone, and therefore has its own currency, whichmeans that there is no risk that Britain will default on its debt. . . . When interestand principal on British government debt come due, the British Government canalways create additional pounds to meet those obligations. By contrast, theFrench government and the French central bank cannot create euros. If investorsare unwilling to finance the French budget deficit—that is, if France cannotborrow to finance that deficit—France will be forced to default. (Russell, 2011).

Prior to the GFC, the validity of the crowding-out argument was rarely questionedby the mainstream financial media. Yet, despite these measured statements byGreenspan and Feldstein, and notwithstanding the empirical evidence, policy-makers continue to insist that financial crowding-out is a significant problem.21

5. Conclusion

The impact of the GFC on deficit and debt ratios in many advanced western econ-omies enables a rigorous assessment of the proposition derived from MMT thatinstitutional arrangements are central to the conduct of fiscal policy. Despitesimilar trends in government debt and deficit indicators, the emerging data hasrevealed differences in the response of long-term interest rates among advancedeconomies. These disparities are most distinct between advanced Eurozone andnon-Eurozone economies.

Mainstream financial crowding-out theory, which is embedded within theneo-liberal expansionary contractions literature and espoused by Inter-Governmental Organisations, albeit with recent qualifications, cannot reconcilethese differences. On the other hand, MMT offers an explanation consistentwith the empirical evidence by making the necessary distinction betweensovereign and non-sovereign economies. However, this has not been recognisedwithin the extant literature.

Panel VECM results suggest that fiscal proxies (cyclically-adjusted financialbalance; the ratio of government gross financial liabilities to nominal GDP; andthe ratio of government net debt interest payments to general government reven-ues) are positive and significant determinants of the long-short interest rate spreadin non-sovereign economies, but are insignificant in sovereign economies.

21Numerous market economists and media commentators have dismissed the claim put forth by JohnBoehner, Speaker of the US House of Representatives, that government spending ‘is crowding outprivate investment and threatening the availability of capital’ (Rowley & Dorning, 2011).


Furthermore, the underlying macroeconomic determinants of the spread for sover-eign economies appear to be different from those of non-sovereign economies.

While market participants seem to have some understanding of the insti-tutional differences between Eurozone and non-Eurozone economies, policy-makers remain wedded to the neo-liberal agenda and fail to make the necessarydistinction for the purposes of conducting fiscal policy. Consequently, sovereigneconomies are not using their inherent (financial) capacity to adopt full employ-ment policies (e.g. a Job Guarantee). Nevertheless, it seems policymakers canno longer use financial crowding-out theory to validate fiscal withdrawal withinsovereign economies.

Acknowledgment

Special thanks to Martin Watts. I am also grateful to an anonymous referee forhelpful comments on an earlier draft.

References

Alesina, A. (2010) Fiscal adjustments: lessons from history. Available at http:// www.economics.harvard.edu/files/faculty/2_Fiscal%20Adjustments_lessons.pdf.

Anderson, T.W. & Hsiao, C. (1982) Formulation and estimation of dynamic models using panel data,Journal of Econometrics, 18, pp. 47–82.

Ardagna, S., Caselli, F. & Lane, T. (2004) Fiscal discipline and the cost of public debt service: someestimates for OECD countries, B.E. Journal of Macroeconomics: Topics in Macroeconomics,7(1), pp. 1–35.

Auerback, M. (2010) The US is NOT Greece, Counterpunch. Available at http://www.counterpunch.org/ 2010/05/07/the-us-is-not-1

Bernheim, B.D. (1989) A neoclassical perspective on budget deficits, Journal of Economic Perspec-tives, 3(2), pp. 55–72.

Bernoth, K., von Hagen, J. & Schuknecht, L. (2004) Sovereign risk premia in the European govern-ment bond market, European Central Bank Working Paper no. 369.

Blanchard, O.J., Chouraqui, J-C., Hagemann, R.P. & Sartor, N. (1990) The sustainability of fiscalpolicy: new answers to an old question, OECD Economic Studies no. 15.

Briotti, M.G. (2005) Economic reactions to public finance consolidation: a survey of the literature,European Central Bank Occasional Paper no. 38.

Buiter, W.H. (2010) The limits to fiscal stimulus, Oxford Review of Economic Policy, 26, pp. 48–70.Caporale, G.M. & Williams, G. (2002) Monetary policy and financial liberalization: the case of

United Kingdom consumption, Journal of Macroeconomics, 23, pp. 177–197.Christopoulos, D.K. & Tsionas, E.G. (2004) Financial development and economic growth: evidence

from panel unit root and cointegration tests, Journal of Development Economics, 73, pp. 55–74.Culbertson, J.M. (1968) Macroeconomic Theory and Stabilization Policy (New York: McGraw-

Hill).De Grauwe, P. (2011) The governance of a fragile Eurozone, Centre for European Policy Studies

Working Paper no. 346.European Commission (2004) Public finances in EMU. Available at http://ec.europa.eu/economy_

finance/publications/publication469_en.pdf.Fullwiler, S.T. (2006) Interest rates and fiscal sustainability, Wartburg College and the Center for

Full Employment and Price Stability Working Paper no. 53.Gale, W. & Orszag, P. (2003) The economic effects of long-term fiscal discipline, Urban-Brookings

Tax Policy Center Discussion Paper no. 8.Galimberti, J.K. & Cupertino, C.M. (2009) Explaining earnings persistence: a threshold autoregres-

sive panel unit root approach, Munich Personal RePEc Archive.

604 T.P. Sharpe

Giavazzi, F. & Pagano, M. (1990) Can severe fiscal contractions be expansionary? Tales of two smallEuropean countries, NBER Macroeconomics Annual, 5, pp. 75–111.

Greenspan, A. (1997) Maintaining financial stability in a global economy, Federal Reserve Bank ofMinneapolis. Available at http://www.minneapolisfed.org/publications_papers/pub_display.cfm?id¼3629.

Gutierrez, M. & Revilla, J.E. (2010) Building countercyclical fiscal policies in Latin America: theinternational experience, World Bank, Southern Africa Poverty Reduction and Economic Man-agement Unit: Africa Region, Policy Research Working Paper no. 5211.

Heller, P.S. (2005) Understanding fiscal space, IMF Fiscal Affairs Department Policy DiscussionPaper no. 4.

Hemming, R., Kell, M. & Mahfouz, S. (2002) The effectiveness of fiscal policy in stimulating econ-omic activity—a review of the literature, IMF Fiscal Affairs Department Working Paperno. 208.

Idier, J., Jardet, C. & de Loubens, A. (2007) Determinants of long-term interest rates in the UnitedStates and the Euro area: a multivariate approach, Banque de France Working Paper no. 170.

IMF (2010a) Moving Public Debt onto a Sustainable Path, IMF Survey Magazine: Policy. Availableat http://www.imf.org/external/pubs/ft/survey/so/2010/pol090110a.htm.

IMF (2010b) Recovery, Risk, and Rebalancing (World Economic Outlook) (Washington, DC: Inter-national Monetary Fund).

IMF (2010c) Fiscal Exit: From Strategy to Implementation (Fiscal Monitor) (Washington, DC:International Monetary Fund).

IMF (2011) Shifting Gears, Tackling Challenges on the Road to Fiscal Adjustment (Fiscal Monitor)(Washington, DC: International Monetary Fund).

IMF (2012a) Coping with High Debt and Sluggish Growth (World Economic Outlook) (Washington,DC: International Monetary Fund).

IMF (2012b) United Kingdom: Article IV consultation, IMF Country Report no. 190.Judson, R.A. & Owen, A.L. (1999) Estimating dynamic panel models: a guide for macroeconomists,

Economic Letters, 65, pp. 9–15.Larch, M. & Turrini, A. (2009) The cyclically-adjusted budget balance in EU fiscal policy making: a

love at first sight turned into a mature relationship, European Commission Economic Papersno. 374.

Lerner, A. (1947) Money as a creature of the state, American Economic Review, 37(2), pp. 312–317.Maddala, G.S. & Wu, S. (1999) A comparative study of unit root tests with panel data and a new

simple test, Oxford Bulletin of Economics and Statistics, 61, pp. 631–652.Marattin, L., Paesani, P. & Salotti, S. (2011) Fiscal shocks, public debt, and long-term interest rate

dynamics, Universita di Bologna Economics Department Working Paper no. 740.Mitchell, W.F. (2009) A modern money perspective on the crisis and a reform agenda, Centre of Full

Employment and Equity Working Paper no. 2.Mitchell, W.F. (2011) MMT—an accounting consistent, operationally-sound theoretical approach.

Available at http://bilbo.economicoutlook.net/blog/?p¼15842.Mitchell, W.F. & Muysken, J. (2008) Full Employment Abandoned: Shifting Sands and Policy Fail-

ures (Aldershot: Edward Elgar).Nickell, S. (1981) Biases in dynamic models with fixed effects, Econometrica, 49(6), pp. 1417–

1426.OECD (2006) Cyclically-adjusted budget balance: a methodological note, OECD. Available at

http://www.oecd.org/dataoecd/0/61/36336878.pdf.OECD (2009) Interim Economic Outlook (Paris: OECD).OECD (2010) Preparing fiscal consolidation, OECD. Available at http://www.oecd.org/dataoecd/16/

1/ 44829122.pdf.OECD (2011) Medium and long-term developments: challenges and risks, OECD Economic Outlook

(Paris: OECD).Paesani, P., Strauch, R. & Kremer, M. (2006) Public debt and long-term interest rates: the case of

Germany, Italy and the USA, ECB Working Paper no. 656.Pesaran, M.H., Shin, Y. & Smith, R.J. (2001) Bounds testing approaches to the analysis of level

relationships, Journal of Applied Econometrics, 16, pp. 289–326.


RBA (2011) Open market operations, Reserve Bank of Australia. Available at http://www.rba.gov.au/ mkt-operations/open-mkt-oper.html.

Rowley, J. & Dorning, M. (2011) Boehner’s views on economy contradicted by indicators, studies.Available at http://www.bloomberg.com/news/2011-05-11/boehner-s-views-on-economy-contradicted-by-indicators-studies.html.

Russell, J. (2011) Martin Feldstein: French ‘don’t get’ problems at euro’s heart, The Telegraph.Available at http://www.telegraph.co.uk/finance/financialcrisis/8981335/Martin-Feldstein-French-dont-get-problems-at-euros-heart.html.

Sharpe, T.P. (2013) Institutional arrangements and public debt threshold limits, International Reviewof Applied Economics, forthcoming.

Sharpe, T.P. & Watts, M.J. (2012) Policy advice in crisis: how inter-governmental organisationshave responded to the GFC, Journal of Australian Political Economy, 69 (Winter), pp.103–133.

Spencer, R.W. & Yohe, W.P. (1970) The ‘crowding out’ of private expenditures by fiscal policyactions, Federal Reserve Bank of St. Louis Review, 52, pp. 12–24.

Sturzenegger, F. (2002) Toolkit for an analysis of debt problems, Centro de Investigacion en Finan-zas Paper no. 12 (prepared for the World Bank).

Wacquant, L. (2001) The penalisation of poverty and the rise of neo-liberalism, European Journalon Criminal Policy and Research, 9, pp. 401–412.

Watts, M.J. & Sharpe, T.P. (2013) Institutional arrangements and the ‘immutable’ laws of debtdynamics, Journal of Post Keynesian Economics, 36(1), pp. 59–84.

Williamson, J. (1990) What Washington means by policy reform, in: J. Williamson (Ed.) LatinAmerican Adjustment: How Much Has Happened? (Washington: Institute for InternationalEconomics), pp. 7–38.

606 T.P. Sharpe

Copyright of Review of Political Economy is the property of Routledge and its content maynot be copied or emailed to multiple sites or posted to a listserv without the copyright holder'sexpress written permission. However, users may print, download, or email articles forindividual use.

sharpe (2014) a modern money perspective on financial crowding out

Documents