MEDIUM RUN EFFECTS OF CONSOLIDATION

consolidation is spending-based and is not completely credible so that figures from meta- studies are used and considering the current crisis situation, in which case multipliers can be expected to be larger and a large part of EU countries would be likely to be in the undesired effect area in the short term.

In any case, it is obvious that this analysis does not suggest that remaining at the baseline, i.e. doing nothing, is preferable to consolidation. In particular, the analysis does not suggest that high- debt countries should consolidate less than low- debt countries, only because they have high levels of debt.

3.2. MEDIUM RUN EFFECTS OF CONSOLIDATION

The previous Section looked at the critical multiplier and presented evidence about how the actual multipliers in EU Member States compare

with the countries' corresponding critical

multipliers showing that it is not impossible that in the current situation consolidation leads to higher debt ratios in the short run. This Subsection looks at how the multipliers affect the debt dynamics following a consolidation, before the next Section introduces possible effects of consolidations on interest rates and moves to look at debt dynamics from a more medium-term point of view.

As will be shown in this Subsection, under most parameter configurations that are in line with the evidence obtained from the economic literature and the simulations presented here, the time period necessary for a consolidation to have a beneficial

effect on the debt ratio is two or maximum three years unless multipliers are very high and persistent. The exception to this pattern occurs in the presence of a high degree of myopia

concerning determinants of government

yields, (108) and this will be examined in the next

Section which will enrich the analysis by including the impact on government interest rates.

As shown in more detail in Box III.3.1 the evolution of the debt ratio, in the absence of any effect on government yields, it is the sum of same three effects indicated in the previous Subsection: i) the cumulative effect of growth on debt, which is generated by the change that the consolidation has

on growth developments. (109) This effect is larger

if the initial debt stock is larger and if multipliers are larger, while simulations show that the impact of the other parameters of the baseline scenario do not have the same relevance. ii) The cumulative effect of growth on government balance which considers how the growth effect of the consolidation affects the budget balance via the operation of the automatic stabilisers on the budget balance. This component of the effect typically increases debt in a consolidation, because it worsens the government balance for a given consolidation and the effect is greater the larger the size of the multipliers and the size of automatic stabilisers. Finally, iii) the cumulative effect from the adjustment of government balance which represents the cumulative savings effect of the consolidation i.e. the direct debt reduction of the adjustment in the absence of any effects on growth. This effect reduces the debt ratio. It increases with the number of years and with the size of the consolidation implemented.

As the first two effects are act to increase the debt ratio, while the third acts to decrease it. One way to look at the medium-term effects of a consolidation, then, is to consider the number of

years n* (hereafter "the critical year") (110)

(108_{) Gros (2011) presents a similar argument in a qualitative} and incomplete manner.

(109_{) The computation takes into account the fact that the} evolution of debt over time is influenced by the path of the government balance, so that a large government balance influences the debt ratio more than a smaller government balance.

(110_{) Notice that n*represents the number of years starting from} the year of consolidation. If consolidation is implemented in year 1, n* represents the critical year. Therefore n*=1

necessary for the consolidation to lead to a decrease in debt with respect to a baseline scenario. In terms of the equation presented in Box III.1 this is equivalent to the number of years

necessary for to equal zero (or be negative).

The critical period n* is different from the number of years required for the debt to go below its starting value in year 0 unless the baseline is the steady state of constant debt ratio. Graph III.3.2 shows illustrative paths for the debt under baseline and consolidation scenarios, for a constant baseline in the left-hand panel and an increasing one in the right-hand panel. It shows that, while in the case of a stable baseline scenario n* coincides with the year in which the debt level returns to its level in the consolidation year, this does not happen when the baseline scenario is increasing and the solid line representing the path of debt-to-GDP ratio following a consolidation returns to the starting level only after crossing the dotted line representing the baseline scenario (if ever). When looking at the effects of a consolidation on the debt, the relevant comparison depends on the aim of the exercise. The debt trajectory under a consolidation should be compared to the baseline debt if we are purely interested in the effect of the consolidation per se; however, if there is an overall

means that there is no self-defeating effect at all, while n*=2 indicates that the perverse effects lasts one year and so on.

question of debt sustainability the debt after a consolidation will also need to be compared to the actual starting level of debt.

In order to model debt dynamics and calculate the value of the critical year n* under different consolidation scenarios, to run debt simulations under different consolidation scenarios, a clear picture of the reaction of GDP to consolidation in

future years (111) is necessary, bearing in mind that

is likely to change over time. The higher the multipliers in the first year and the longer the change in GDP induced by the consolidation, the larger the value of n* and the longer it will take for a consolidation to be effective. There are three broad effects emerging from the literature review presented in Chapter III.2 that will be incorporated into the simulations presented later in this Subsection:

i) the values of first year multipliers in normal economic conditions are typically estimated at between 0.3 to around 1 depending on the composition and nature of the policy changes and type of estimate; ii) the values of first year multipliers are larger in crisis years, usually within a range from 0.5 to 1.6; and iii) the impact on GDP in years following consolidation tends to decrease but does so to different degrees depending on the

(111) See Box III.1 that defines the adjusted multiplier which corresponds to the difference between GDP and baseline at year t in an impulse-response function.

Graph III.3.2: Critical year and underlying debt trend

92% 94% 96% 98% 100% 102% 104% 0 1 2 3 4 5 6 7 8 year

baseline scenario deviation

94% 96% 98% 100% 102% 104% 106% 108% 110% 112% 0 1 2 3 4 5 6 7 8 year

baseline scenario deviation

model used, on the type of estimate and on all the factors affecting multipliers.

The fiscal multipliers can be very persistent or can decay rapidly in the first years, following a convex, autoregressive path. Such an AR1-shaped curve is similar to the shape of GDP responses which can be found in New Keynesian DSGE- based assessments of multipliers for various (but

not all) types of consolidation. (112) Graph III.3.3

shows two stylised GDP responses following a consolidation of 1% of GDP, under low and high

persistence. (113) The main difference between the

two paths thus concerns the persistence of the effects of the consolidation.

Graph III.3.3: Stylised paths of GDP impulse responses used in the simulations 0 0.5 1 1.5 2 2.5 1 2 3 4 5 6 7 8 9 10 G DP Res po ns e Years HP LP

Source: SCPs, Commission services.

Simulations

Graph III.3.4 shows the debt-to-GDP ratio dynamics for the low-persistence multipliers path under different assumptions about the impact multiplier. The baseline scenario is one of a constant debt ratio of 100% of GDP. The Graph shows debt dynamics for a persistence rate of

0.5, (114) with first year multipliers of 0.5, 1 and

(112_{) See for example in Graphs III.6.1 and III.6.2 in European} Commission (2010a).

(113_{) The precise formula is shown in Box III.1 as equation III.6.} It is considered that the multipliers decline according to a persistence parameter to reach a long-term value. The persistence parameter is the ratio between the responses of two consecutive years if the long-run impact of fiscal consolidation is null.

(114_{) 0.6/0.7 is the ratio of second to first year GDP responses in} the case of composition-balanced permanent consolidation

1.5. All values for the first-year multiplier that lie below the 0.7 level will correspond to an improved debt ratio from the first year – this is so by construction as the 0.7 level corresponds to the critical value for the multiplier. It should be noted that a first year multiplier of 1.5 is on the high side of existing estimates as it is the estimate of a temporary consolidation based on government spending.

Graph III.3.4: Debt dynamics (baseline steady state, b0 = 100%), no effect on interest rates – with low persistence

93% 94% 95% 96% 97% 98% 99% 100% 101% 102% 0 1 2 3 4 5 6 year m = 0.5 m = 1 m = 1,5 baseline

Source: Commission services.

Graph III.3.5 shows the case for a high persistence parameter (0.8) of the GDP response. The higher persistence of the effects of consolidation generates longer-lasting negative effects from fiscal consolidation. If the first-year multiplier is 1.5 the consolidation-based debt increase lasts for one more year so that three years are needed – taking into account the fact that year 1 is the year in which the consolidation is implemented – before debt goes below baseline.

in European Commission (2010a). This is the basis for the choice of 0.5 as low persistence and 0.8 as high persistence parameters. Note that the persistence in the following years is however smaller. Values of the GDP responses broadly constant for the first three years are very commonly found in VAR estimates. This wold make raise an hump-shaped GDP response with the consequence that the debt increases following a consolidation would be reversed only after three years for values of the impact multiplier of 1.5. This being the only difference, the case is not developed here.

Graph III.3.5: Debt dynamics (baseline steady state, b0 = 100%), no effect on interest rates – with high persistence

93% 94% 95% 96% 97% 98% 99% 100% 101% 102% 0 1 2 3 4 5 6 year m = 0.5 m = 1 m = 1,5 baseline

Source: Commission services.