One of the most important questions in the current process of fiscal consolidation in many developed economies concerns the size and the pace of the adjustment. An excessive and/or too-fast fiscal retrenchment can have dramatic effects on unemployment and growth, while if it is too slow, it can prove to be ineffective and lack credibility in the eyes of the financial markets. Thus, when the debt-to-GDP ratio is high and there is limited fiscal space, the challenge is to find the proper balance between growth, efficiency and credibility of the fiscal adjustment.
One of the main factors that determine this credibility is whether the implementation of the fiscal adjustment brings about an effective reduction in the structural budget balance, i.e., the public deficit that exists once the effect of the economic cycle has been adjusted for. The European Commission has just published its winter forecasts for 2013, where it has updated its estimates on structural budget balance. According to these estimates, for ten of European economies the structural public deficit exceeds 3% of their GDP. Spain is the country with the third highest structural deficit in 2012, after Ireland and the UK, with 5.9% of GDP.1
The budget balance and the unemployment rate
Such a high estimated structural deficit for the Spanish economy is particularly striking considering that the unemployment rate was 25% in 2012, which is prima facie evidence of a strong negative cyclical position. As the public deficit converges towards its structural level (when the economy tends towards a neutral cyclical situation), unemployment simultaneously converges towards its structural rate. However, neither of these structural components can be observed and have to be estimated, introducing uncertainty in key variables for economic policy.
Figure 1 represents the Spanish budget balance, as a percentage of GDP, against the unemployment rate, which appears on the horizontal axis; the negative correlation among these variables stems from the effects of the automatic stabilisers.2 Figure 1 also displays two vertical dashed red lines corresponding to the structural unemployment rates in 2006 and 2012 estimated by the Commission.3 According to these estimates, structural unemployment appears to be highly volatile and strongly procyclical ranging from 11% (when the current unemployment rate was 8.5%) up to 21.7% (25% respectively), almost eleven points of difference in only six years for a structural component that should be otherwise fairly stable. Thus, these estimates imply that most of the fall in economic activity in Spain in recent years has meant a fall in potential GDP. To put it another way, that practically two thirds (0.65 of a percentage points) of each point of increase in the unemployment rate can be considered as structural.
Figure 1. Spain, general government budget balance and unemployment rate, 1980-2012 (source: European Commission and own elaboration)
On top of these vertical dashed red lines for the structural unemployment rates for 2006 and 2012, the chart displays the structural budget balance estimated by the Commission for those two years; 1.7% and -5.9% respectively. Given that the nominal budget balance in 2006 was 2.4% of GDP, we can infer that the cyclical balance estimated at the highest point of the real-estate boom reached barely 0.7 percentage points of GDP. This is contradictory with the estimates by some authors (for example, Martínez-Mongay, Maza and Yaniz 2007), who found that between two and three points of GDP could be considered to be transitory revenues in those years.
According to the decomposition estimated by the Commission, in 2012 for each point of cyclical unemployment the budget balance fell by 0.63 of a percentage point; this value is the result of dividing the cyclical fiscal balance (that is, -2.1% is equal to -7%, -1% and 5.9%), including one-offs (1%), by the cyclical unemployment rate (that is 3.3% is equal to 25% minus 21.7%). At least for 2012, the sensitivity of the cyclical component of the budget balance to the unemployment rate estimated by the Commission is very close to the value obtained by directly regressing the total budget balance on the unemployment rate (the regression coefficient is equal to -0.68).4
Therefore, Figure 1 makes clear that the structural budget balance estimated by the Commission depends on two crucial factors:
1) The estimated structural unemployment rate.
2) The sensitivity of the budget balance to cyclical unemployment.
An alternative estimate of the structural budget balance
How sensitive is the structural budget balance to these two factors? To answer this question we introduce two changes with important implications:
- The structural unemployment rate is estimated following an alternative procedure base on the Okun's Law (see, for example, Ball, Leigh and Loungani, 2013), according to which this rate was 14% in 2006 and 18% in 2012, less volatile and procyclical than the one estimated by the Commission. Consequently cyclical unemployment is more persistent.5
- The sensitivity of the cyclical budget balance to cyclical unemployment has been re-estimated, giving a ratio of -0.7, very similar to the regression coefficient of the current budget balance over the unemployment rate (-0.68) and slightly above the elasticity estimated by the Commission (-0.63).6
With these new estimates the structural budget balance in 2012 can be calculated projecting the observed deficit onto the new vertical solid green line in the figure: 3%. Likewise, the structural budget balance in 2006 would have been in deficit (-1.5%), even though there was a positive balance due to the asset-boom effect, largely on revenues. The comparison between the results from these two methods highlights the paramount importance of uncertainties in both, the estimate of structural unemployment (e.g. Staiger, Stock, and Watson 1997), similar to that existing with trend GDP (Larch and Turrini 2009), and in the sensitivity of the deficit to the economic cycle. The question of which method to use, however, may be based on diverse qualitative judgements:
- For 2006 the alternative approximation provides results that are more consistent with the existence of an asset boom.
Between 2007 and 2008 non-financial public revenue fell 4.1 points as a percentage of GDP, while the structural unemployment rate estimated by the European Commission would have risen by only 1.4 points, from 11.8 to 13.2%:
- The Spanish economy is implementing reforms (in the labour market and in the goods and services markets) that should reduce structural unemployment and, therefore, the structural deficit.
- A reasonably high number of fiscal-adjustment measures in 2010, 2011 and, particularly, in 2012, can be considered genuinely structural.
Economic policy implications
According to our calculations, the Spanish economy would still need three points of GDP fiscal adjustments in the years to come to reach the budget equilibrium mandatory in the new Spanish Budget Stability and Financial Stability Act. With the calculations of the Commission the required adjustment would have to be almost twice as high.
Given the uncertainty regarding the estimation of structural budget balances and unemployment, and about the negative effects of fiscal adjustments on economic activity, it is advisable to be cautious and not carry out fiscal adjustments that could end up being excessive, at least until further information allows a more precise estimate of the structural deficit. The implication of these discrepancies for this year’s fiscal objective is straightforward. As Buti and Carnot (2013) remind us, the IMF (2012) considers a structural adjustment pace of 1% a year as a useful guideline, and the SGP requires an annual structural adjustment of 0.5% or more in the case of vulnerable countries. The problem is how to make this rule operational in a new fiscal target for 2013, given the uncertainty surrounding the structural unemployment rate. Assuming that the unemployment rate will increase from 25% to 26,9% (as in the recent winter forecasts by the Commission), the question is by how much is the structural unemployment affected by this increase in the observed rate. In fact, depending on the assumptions about the increase of the structural unemployment rate (and given an initial budget deficit of 6.7% in 2012), a fiscal target ranging from -6.1% to -6.7% in 2013 could be sufficient to ensure a 1% structural adjustment this year.7
Given these uncertainties, 2013 could be a year of transition in which it would suffice to apply the fiscal measures already under way, working rather on the quality of the fiscal adjustment. For example, substituting the one-off measures in 2012 by permanent ones and implementing much more selective measures, which might allow improvements in the efficiency of public administrations, based on items with lesser short term impact on growth and wellbeing.
Meanwhile, it is necessary to continue to apply new structural reforms and also to give time to those that are already under way to do their job. As a result of these reforms, for each point of reduction in structural unemployment, we reduce the structural deficit by 0.7 percentage points of GDP. In short, the increase in potential growth, the reduction of structural unemployment and the fiscal consolidation are processes that mutually reinforce each other.
Ball, L, D Leigh and P Loungani (2013), “Okun’s Law: Fit at 50?”, NBER Working Paper 18668.
Buti, M and N Carnot (2013), “Fiscal policy in Europe: Searching for the right balance”, VoxEU.org, 14 March.
Corrales, F, Doménech, R and J Varela (2002), "El Saldo Presupuestario Cíclico y Estructural de la Economía Española", Hacienda Pública Española, 1(3), 1-26.
Denis C, Grenouilleau D, McMorrow K and W Roeger (2006), “Calculating Potential Growth Rates and Output Gaps – A Revised Production Function Approach”, Economic Papers 247, European Commission.
Fedelino, A, A Ivanova and M Horton (2009), “Computing Cyclically Adjusted Balances and Automatic Stabilizers”, Technical Notes and Manuals, IMF.
IMF (2012), “Nurturing credibility while managing risks to growth”, Fiscal Monitor Update, July.
Larch, M, and A Turrini (2009), “The cyclically-adjusted Budget balance in EU fiscal policy making: A love at first sight turned into a mature relationship”, Economic Papers 374, European Commission.
Martínez-Mongay, C, L A Maza Lasierra and J Yaniz Igal (2007), “Asset Booms and Tax Receipts: The case of Spain, 1995-200”, European Economy Economic Papers 293.
Mourre, G, G M Isbasoiu, D Paternoster and M Salto (2013), “The cyclically-adjusted budget balance used in the EU fiscal framework: an update”, Economic Papers 478, European Commission.
Staiger, D, J H Stock and M W Watson (1997), “How Precise Are Estimates of the Natural Rate of Unemployment?”, in C D Romer and D H Romer (eds.), Reducing Inflation: Motivation and Strategy, University of Chicago Press, 195–246.
1 Given that the deficit for 2012 was 7% and that the one-offs measures estimated by the Commission represent 1% of GDP, this means that the cyclical component of the public deficit only accounts for 2.1 percentage points of GDP. In other words, according to the Commission most of the deficit in 2012 is largely due to its structural component.
2 The correlation between the budget balance and the unemployment rate was equal to -0.84 between 1980 and 2012.
3 See Denis et al (2006) for a description of the method used to estimate the structural unemployment rate.
4 This was not the case in 2006, when the cyclical budget balance was 0.8% and the cyclical employment -2,46%. The slope for that year (-0.33) was half of the slope estimated in 2012, and for the period 1995-2012 has been equal to -0.456. The changes in these elasticities are explained by the fact that the in the method implemented by the European Commission, the sensitivity of the deficit to the economic cycle depends on the elasticity of revenues to their respective tax bases, and of the latter to the output gap and not to the unemployment rate (see, Fedelino, Ivanova and Horton, 2009, and Mourre et al, 2013).
5 It could be argued that many jobs that have been destroyed are of a permanent nature (for example, in the construction sector). Unless there is a complete hysteresis, the permanent destruction of jobs in certain specific sectors does not necessarily mean that the workers who held these jobs will become permanently unemployed, thus prompting an increase in the structural unemployment rate.
6 To obtain the elasticity of the cyclical budget balance to the cyclical unemployment rate we follow the procedure described by Corrales, Doménech and Varela (2002), but using the unemployment rate instead than the GDP.
7 One possibility is to assume that structural unemployment rate in 2013 will increase in the same proportion as the one estimated by the Commission from 2006 to 2012 (65% of the increase of the unemployment rate was due to its structural component). On the contrary, according to our alternative estimate, the structural unemployment rate increased only 4 pp out of the 16,5 points observed in the current unemployment rate, that is, only 24 per cent of the total increase would be explained by its structural component.. The implications of these different assumptions in terms of the expected increase of the cyclical unemployment rate in 2013 are huge (0.67 points vs 1.44). Given that the preliminary estimate of the budget balance in 2012 was -6.7%, first we could add between -0.4 and -1.0% of expected cyclical budget balance in 2013 (the elasticity, -0.63 vs -0.70, times the increase of the cyclical unemployment rate, 0.67 vs 1.44). Second, we subtract 1% of structural fiscal consolidation. Under these assumptions, the 2013 fiscal target would range from -6.1% to -6.7%.