The Changing Value OfMultipliers and Fiscal Policy

Of special interest is a critical examination of two positions that have inspired recent fiscal policy attitudes against the crisis in Europe and the idea of limited effectiveness of fiscal policy.

As Gali, Lopez-Salido and Valles (2007) underline, the New Keynesian model does display non-Keynesian effects, unless liquidity-constrained agents are present, as their con­sumption will increase with real wages and other factors setting a constraint on their budgets. In reality, a significant fraction of consumers is liquidity constrained (see e.g. Di Bartolomeo, Rossi and Tancioni 2011), and this probably increases in recessions, which leads to time-varying fiscal multipliers. This, together with sticky prices, justifies the pursuit of countercyclical fiscal policies, with government spending set at a level above that warranted by the ‘efficient provision of public goods' (Gali 2005: 587-88).

A number of factors can influence the effectiveness of fiscal policy. We are interested in knowing whether it increases in downturns relative to ‘normal' circumstances and how the fiscal multipliers vary across different fiscal instruments or with other policy instruments in use at the same time and with market expectations of future events or with the size of likely shocks. The lesson is that these are all elements crucial to understanding the potential effects of fiscal policy, in particular, the sign and the size of the different fiscal multi­pliers, and their interaction with, and dependence on, the monetary policy stance.

Blanchard, Dell'Ariccia and Mauro (2010) draw a number of lessons from the crisis. First, the Great Moderation was useful as a way to stabilise expectations in an environment of supply shocks (e.g. for oil prices) but has possibly nourished one of the sources of the Great Recession, i.e. neglect of risks by the agents. Second, financial intermediation matters, but its role can be impaired by a number of circumstances, in particular (as we will see below) as far as the transmission of monetary impulses from the short run to the long run is concerned. In addition, regulation is not macroeconomically neutral, as shown by the crises derived from the great bubble bursts in 2001 and 2007-8. Finally, countercyclical fiscal policy matters for a number of reasons: (1) because monetary policy has well-known limits of effectiveness, especially when the ZLB is approached and (2) because it has longer lags in its effects (even if lower implementation lags), which is in contrast with the longer administrative lag required for introducing fiscal policy. The total lag is shorter for fiscal policy, which has thus returned to centre stage, also due to the prospective length of the crisis.

Other studies stress the efficacy of fiscal policy in severely depressed economies when central banks do not offset its effects (DeLong and Summers 2012). Moreover, fiscal multipliers are shown to be asymmetric and regime dependent (Auerbach and Gorodnichenko 2012; Mittnik and Semmler 2012). They are ‘stronger in recessions than in expansions, in particular in presence of financial market stress, so that contractionary effects can become very severe when fiscal consolidations are pursued' (Semmler and Semmler 2013: 2), as an effect either of some economies being locked in a bad equilibrium (De Grauwe 2011) or of macroeconomic non­linearities (Semmler and Semmler 2013). This result is confirmed by the analysis of seven structural DSGE models used for policy action as well as academic DSGE models (see, in particular, Coenen et al.

These findings must be assessed in conjunction with those of Bilbiie, Monacelli and Perotti (2014), according to which usually government expenditures at the ZLB do not necessarily enhance welfare, even when their output multi­plier is large. However, in another paper, Bilbiie, Monacelli and Perotti (2012) show that tax cuts financed with govern­ment debt at the ZLB are Pareto improving when there are constrained borrowers, as the cuts are a form of implicit transfer from unconstrained savers to constrained borrowers.

Of specific relevance are some analyses that take account of open economies (in some cases the EMU) and spill-over effects.

In summary, the size of multipliers varies according to a number of circumstances: the type of fiscal instrument used for consolidation, the cyclical situation of the economy, the expansionary or contractionary nature of other policies (notably of monetary policy), the degree of price and wage flexibility, the degree of openness of the economy and the policies enacted in related economies as well as their rates of growth.

Let us turn now to the second issue mentioned at the beginning of this section, i.e. the need for fiscal consolida­tions. This can arise from different considerations. The first consideration is that accumulation of debt aggravates the fiscal burden of future generations. In addition, it could have negative effects on various other relevant policy targets.

In particular, it could reduce growth. The assertion about the beneficial effects of fiscal austerity has passed through a long process of theoretical refinements, confutations and empirical evaluations. Recently, the negative effects of debt on growth have received theoretical support by Checherita- Westphal et al. (2014). However, doubt with respect to some tenets of this idea was first raised by Blanchard and Perotti (2002), who gave a substantially Keynesian answer to the issue of the effects on income of the tax increases and expen­diture cuts needed for consolidation. In fact, they found that the former have a contractionary effect, whereas the latter have an expansionary one. The authors did not engage in a discussion about debt-consolidation strategies, but on the basis of their findings, one could hardly assert that a policy of expenditure reductions and (to a less extent) tax increases, while certainly contributing to the reduction of the numera­tor of the debt/GDP ratio, would give an impulse to the denominator. Instead, their findings might support debt con­solidation not based on a budget contraction, at least as far as the effects on income are concerned. In addition, with spe­cific reference to the effects on growth of one item of public expenditures, i.e.

As far as empirical tests are concerned, on the basis of a long time series of data referred to a wide range of countries, Reinhart and Rogoff (2010) draw the conclusion that the relationship between debt and growth is weak for debt/GDP values below 90 per cent but becomes significant and nega­tive beyond that level. Herndon, Ash and Pollin (2013) have criticised these results due to the omission from the Reinhart and Rogoff's test of important developed nations - which experienced high growth even with a high debt burden - and low weights on countries for the duration of high debt and growth performance. When account is taken of this, the 90 per cent value is no longer a ceiling beyond which growth is reduced.

A rather complete and detailed - unfortunately, not updated to the wealth of more recent contributions - empirical analysis of the effects of fiscal consolidation is presented in IMF (2010), which takes account of numerous aspects of the effects of fiscal consolidation policies: in par­ticular, their timing (i.e. whether they are short or long term), the monetary policy stance and the expansionary or contrac­tionary nature of budget policies in other countries. Its con­clusion is that, first, the idea that fiscal austerity triggers faster growth in the short term finds little support in the data. Typically, fiscal retrenchment has contractionary short-term effects on economic activity, with lower output and higher unemployment, but ‘fiscal consolidation is likely to be beneficial over the long term' (IMF 2010: 113). In addition, a budget cut is less expansionary the lower the interest rate (as monetary policy has little room for partially accommodating the deflationary effects), the lower the like­lihood of a currency depreciation and the less expansionary are the policies of other countries, which gives little scope for raising net export. Also critical towards the expansionary effects of fiscal consolidation is Perotti (2013), who shows that at least in the three episodes referred to by Giavazzi and Pagano (1990, 1996) in their papers on the expansionary virtues of fiscal consolidation, those of Ireland, Denmark and Sweden - the expansion derived from either external demand or by other factors. Similar results hold for consoli­dations implemented by Belgium, Finland and the United Kingdom. An important ingredient of the success of consoli­dation was incomes policy, even if it offered only temporary beneficial support.

Some authors point out that smoother fiscal consolidations are more successful than stiffer ones (Batini, Callegari and Melina 2012). In addition, consolidation is more successful if policies are well coordinated, the growth rate is positive, and tax yields or welfare spending is not too high a share of GDP. When the monetary policy stance is accommodative - and the longer so - multipliers rise. Coenen et al. (2012) and Gechert, Hughes Hallett and Rannenberg (2015) have esti­mated that if this lasts, the value of cumulative spending multipliers rises well above 1. Similar is the effect of the private sector's lasting expectations about an accommoda­tive monetary policy. Finally, the degree of openness of the economy and the rate of growth of other economies are rele­vant due, as indicated earlier, to the fact that multipliers are lower for an open economy, the more so the higher is its degree of openness and the size of demand abroad. Moreover, consolidation is more likely to lead to success when it takes place in a small economy, out of pre-election years and where governments face coherent oppositions (Fatas et al. 2003).

Anyway, Blanchard and Leigh (2013a) find that stron­ger planned fiscal consolidations were associated with lower growth in developed economies. Thus, Blanchard and Leigh's advice (2013b) is to proceed very carefully in deciding when to proceed to choices of consolidation and how much consolidation to implement. An impor­tant variable is political, in that deferring consolidation can require consideration of political factors at work. Consolidation, in fact, may need to be continued for a long time, spanning the length of a government's mandate.

These findings have important consequences for our topic, as far as both institutional and short-run implications are concerned. We will briefly deal with them in the next section.

3.9