Rebalancing in Spain’s private sector is under way, but with more modest progress on reducing stocks. Spain is subject to significant spending pressures, reflecting unfavorable demographic trends and subdued growth prospects, and will require substantial structural reform. Priority reforms are needed to strengthen its fiscal framework. The study tries to infer the potential impact of the ongoing integration process on bank efficiency based on preconsolidation bank data. The reforms, such as those recently implemented, of employment protection and collective bargaining could help improve Spain’s inflation performance.

Abstract

Rebalancing in Spain’s private sector is under way, but with more modest progress on reducing stocks. Spain is subject to significant spending pressures, reflecting unfavorable demographic trends and subdued growth prospects, and will require substantial structural reform. Priority reforms are needed to strengthen its fiscal framework. The study tries to infer the potential impact of the ongoing integration process on bank efficiency based on preconsolidation bank data. The reforms, such as those recently implemented, of employment protection and collective bargaining could help improve Spain’s inflation performance.

VI. Determinants of Spanish Inflation: the Role of Labor and Product Market Institutions1

Spain needs lower inflation than elsewhere in the euro area. But its track record is poor. Econometric evidence points to a large role of Spain’s labor market institutions in explaining this poor performance, especially the intermediate collective bargaining system and the high degree of employment protection. These results suggest that the reforms, such as those recently, of employment protection and collective bargaining could help improve Spain’s inflation performance and contribute to achieve the needed gains in competitiveness.

A. Motivation

1. Spain needs lower price growth than its partners. In order to regain competitiveness and strengthen external sustainability without excessive reduction in absorption, Spain needs to keep its inflation below that of its euro area partners for some years. But since the launch of the euro, inflation has tended to be above the euro area average, leading to a cumulative price differential of about 10 percent (Figure 1). And while during the crisis the inflation differential turned briefly negative, it has now returned to positive values, despite the greater slack in economic activity in Spain than in the euro area as a whole. And while the recent high inflation reflects temporary factors, there is a risk that it may lead to second-round effects.

Figure 1.
Figure 1.

Spain and Euro Area: Headline Inflation

(year-on-year percent change)

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A006

Source: Eurostat.

2. This paper examines the role of inefficiencies in labor and product markets in keeping inflation high in Spain. Inflation differentials are per se not always worrisome. Some inflation differentials are benign, either because they reflect a catch-up process, are part of an equilibrating mechanism, or result from temporary shocks. However, inefficiencies in domestic product, labor or other factor markets can amplify or make more persistent the impact of shocks on inflation. The paper estimates a model of the determination of inflation in ten euro area countries to measure the contribution of these various factors. Based on this analysis, it discusses which policies could be implemented in Spain to avoid the recurrence of a persistent inflation differential in periods of boom and to facilitate the currently ongoing process of internal devaluation.

B. Stylized Facts

The boom years

3. Spain has maintained one of the largest inflation differentials with the euro area for most of the period 1999–2008. The differential averaged 1 percentage point over this period, and was only slightly surpassed by Greece and Ireland. Portugal and Luxembourg had the next largest differentials at 0.7 percentage points (Table 1).2 The GDP deflator and core inflation show a similar picture.3 In addition to stronger core inflation, inflation in Spain was also pushed above the euro area average by a larger contribution of food price inflation (reflecting both a higher share of food in the consumption basket and stronger food price increases) (Figure 2, Figure 3). The contribution of energy prices was not different from the euro area average.

Table 1.

Average Inflation Differentials with Respect to Euro Area, 1999-2008 1/2/

article image
Source: Eurostat.

2001-2008 for Greece’s GDP deflator inflation.

EA11-16 for headline and core inflation; EA16 for GDP deflator inflation.

Figure 2.
Figure 2.

Spain: Contributions of Inflation Components to Inflation Differential with Euro Area

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A006

Source: Eurostat.
Figure 3.
Figure 3.

Spain: Accumulated Price Differences Relative to Euro Area

(index, 1999=100)

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A006

Source: Eurostat.

4. Price increases were strongest in the services sector and in construction. Consumer price inflation in services was much stronger than in the euro area, while the differential with the euro area for inflation in goods was smaller, whether one looks at producer price inflation or prices of goods after retail trade (see Figure 3). Services in the consumption basket include items such as communication, housing, recreation and personal care, and transport. Looking at the supply side, increases in value-added prices (basic prices) were stronger than in the euro area in construction, trade/transport/communication, and to a lesser extent in industry (Figure 4).

Figure 4.
Figure 4.

Spain: Inflation Differential with Euro Area by Value Added Sector

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A006

Source: Eurostat.

5. Higher Spanish inflation reflected stronger increases in both unit labor costs (ULC) and unit gross operating surplus (UGOS). GDP deflator inflation can be decomposed into the contributions of unit labor costs, the unit gross operating surplus and unit net taxes (taxes net of subsidies). Up to 2008, ULC and UGOS contributed about equally to the inflation differential. ULC and UGOS each grew by about 15 percentage points more than in the euro area between 2000 and 2008 (see Figure 2, Figure 3). Interestingly, the ULC growth differential with the euro area was exclusively due to faster growth in labor cost per hour (rather than slower labor productivity growth per hour) (Table 2).

Table 2.

Contributions of Labor Cost and Labor Productivity to Nominal ULC Growth

article image
Source: Eurostat.

6. The sectors with high ULC growth and/or high return on capital were construction and services sectors.4 5 Construction, primary activities, health/social work, and trade had both strong ULC growth and rate of return on capital (Table 3). Real estate activities and transport and storage had relatively high ULC growth. Financial intermediation, hotels and restaurants had relatively high rates of return on capital.

Table 3.

Spain: Sectors with High Unit Labor Cost Growth and/or High Rate of Return on Capital

article image
Sources: Eurostat; and EU Klems Database.

Relative to average for Spanish economy and relative to Germany.

The Great Recession

7. Inflation has moderated during the Great Recession, although not enough to start correcting substantially the accumulated inflation differential with the euro area. Spanish inflation turned slightly negative in 2009 but returned to 2 percent in 2010. The differential with the euro area which had also turned slightly negative in 2009 returned to positive territory in 2010 due to transitory factors, in particular larger energy price increases in Spain and the increase in VAT. The inflation differential excluding changes in indirect taxes remained favorable to Spain in 2010 and early 2011 (see Figure 1). Core inflation also remained moderate relative to the euro area. This adjustment, however, is small relative to the accumulated inflation differential and no significant correction has been achieved yet (Figure 5). It is also small relative to that of Ireland and, to a lesser extent, Portugal.

Figure 5.
Figure 5.

Selected Euro Area Countries: Accumulated Price Differences Relative to Euro Area, 1999-2010

(index)

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A006

Source: Eurostat.

8. The adjustment during the Great Recession was mostly borne by ULC, as employment fell drastically. ULC moderated significantly, even turning negative in 2010. They grew 3 percentage points less than those of the euro area over 2008-2010 (see Figure 2). This was due to increases in labor productivity growth, which reflected Spain’s particularly sharp fall in employment, rather than stronger labor cost moderation (see Table 2). In contrast, UGOS moderated much later and less than ULC and than the UGOS of the euro area, contributing to offset the impact of the moderation of ULC on inflation. Net taxes also contributed to push the inflation differential with the euro area over these years.

C. Analysis

9. A number of factors contribute to determine inflation, some of which are not worrisome. Some inflation differentials are benign, either because they reflect a catch-up process, are part of an equilibrating mechanism, or result from temporary shocks.6

  • First, inflation could be temporarily higher in some countries due to a catch-up from initially low price levels. The adoption of a common currency has increased market integration and price transparency, reducing the scope for deviations from the law of one price. Some studies (e.g. Honohan and Lane, 2003) find that this factor explained a large part of the inflation differentials in the euro area at the beginning of EMU.

  • A second benign explanation often referred to is the Balassa-Samuelson effect. Countries that are in the process of catching up in living standards usually experience stronger productivity growth in the tradable sector. This leads to wage increases in the tradable sector which, if labor mobility between sectors is high, also increases wages in the nontradable sector. Given that productivity growth is usually lower in that sector, it pushes prices up in the nontradable sector, contributing to higher inflation. Rabanal (2009) finds little evidence of such an effect in Spain, while Beck et al. (2009) and ECB (2005) also find little support for such a relationship across broader samples of euro area countries.7

  • A third explanation for inflation differentials is differences in business cycles. The output gap has been found to be a significant determinant of inflation (e.g. Honohan and Lane, 2003; Andersson et al, 2009), with a higher output gap leading to higher inflation. In this case, the higher inflation may be part of the equilibrating mechanism, though it may take time to operate in a monetary union.8

  • Fourth, asymmetric supply (e.g. labor productivity shock) and demand shocks (e.g. fiscal policy shock) will lead to inflation differentials in a monetary union and these are part of the equilibrating mechanism. Common shocks to the monetary union could also lead to inflation differentials, due to different economic structures in the various countries. For instance, one common shock that has received a lot of attention is a change in the euro exchange rate. The impact on domestic inflation will differ depending on the share of consumption goods imported from outside the monetary union and indirectly, through the competitiveness channel, on the degree of openness to countries outside the monetary union. Studies in the literature have yielded conflicting evidence on the role of differences in nominal effective exchange rates, with the later studies concluding the impact is minor (see Honohan and Lane, 2003 and 2004 versus Angeloni and Ehrmann (2007) and Andersson et al. (2009)). Different specialization of the economies could also lead to an asymmetric impact of common shocks.

10. In contrast, sources of inflation differentials that could lead to undesirable outcomes are structural inefficiencies in domestic product, labor or other factor markets.9 Some institutions may amplify or make more persistent the impact of shocks. For instance, intermediate coordination in wage bargaining could lead to a less efficient response of inflation to supply shocks than highly coordinated or fully decentralized systems. Indeed, in bargaining systems with intermediate coordination, unions can exert some market power on wage setting but tend to ignore the macroeconomic implications of their actions (e.g. Calmfors and Driffill, 1988). In contrast, in fully coordinated systems, unions recognize their market power and take into account the impact of their wage demands on inflation and unemployment. In fully decentralized systems, unions have very limited market power. Hence, second-round effects of a supply shock (e.g. oil price increase) would be stronger and lead to higher inflation and unemployment in an intermediate system of bargaining.

11. Looking initially at where Spain stands on the various determinants, several factors could explain why its inflation was higher than the euro area average over 1999-2008.10 These include a relatively low initial price level (at about 87% of the euro area average in 1999) and an average output gap over the period significantly larger than in most other euro area countries (Figure 6). In addition, while employment protection and collective bargaining have been recently reformed, up to 2009 Spain’s labor market was characterized by less efficient institutions, especially an intermediate coordination of collective bargaining and relatively high EPL (though this was partly offset by a low union density). In addition, there is a high degree of inflation indexation, which amplifies second-round effects of supply shocks, causes a high wage drift and reduces the sensitivity of real wages to the economic cycle. In contrast, Spain performs relatively well on product market regulation, at least as far as the overall indicator is concerned. The available indicators of product market regulation, however, miss very important elements, such as the construction sector that was crucial to explain the behavior of mark-ups in the Spanish economy, as identified above.11

Figure 6.
Figure 6.

Spain and Euro Area: Determinants of Inflation

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A006

Sources: Eurostat; and OECD.

12. The paper examines the role of labor and product market institutions in explaining inflation developments. In line with Bowdler and Nunziata (2007), Biroli et al. (2010) and Correa-López et al. (2010), a traditional backward-looking Phillips curve equation is augmented with structural labor and product market indicators. In this model, inflation is a function of its own lag (to capture persistence), the initial relative price level, the output gap, changes in the nominal effective exchange rate and common shocks (e.g. oil price shock, monetary policy shock captured by time dummies).12 The model is augmented to look at the impact of four institutions, namely the degree of coordination of collective bargaining, union density (i.e. the number of employees registered with unions), employment protection, and product market regulation. The extent of inflation indexation could not be introduced, due to the lack of a broadly available indicator. Following other studies, these factors are allowed to affect inflation through their impact on (i) inflation persistence; (ii) the response of inflation to the output gap; and (iii) the response of inflation to common shocks.13

13. The model is not without potential caveats and these will be addressed in the robustness tests. First, survey evidence for euro area countries suggests that inflation may not depend only on lagged inflation but also on future expected inflation, as when prices are sticky, profit-maximizing firms take into account expected future developments to set prices. Second, the analysis assumes that cross-country heterogeneity in the responsiveness of inflation to its determinants is fully accounted for by differences in the considered labor and product market variables. This common elasticity assumption could be too stringent.

14. The model is estimated over a panel dataset of 10 euro area countries over the period 1983-2007.14 Several estimation methods are used, including simple OLS, estimation with country fixed effects, and instrumental variable estimation (with country fixed effects) to correct for the potential endogeneity of the output gap and of the lagged inflation variables.15 Outliers are excluded based on statistical tests; standard errors are robust; and structural variables are standardized. In a first step, we estimate a linear version of the model that excludes the interactions between common shocks and structural variables (Table 4). In a second step, we estimate the full model including the interactions between common shocks and structural variables (Table 5). Finally, specifications are reduced to eliminate insignificant variables through an iterative process in which the least significant variable is eliminated and the model is re-estimated until all variables are significant at least at the 10% level. Several results emerge from the estimations.

Table 4.

Determinants of Inflation

(linear model)

article image
Source: IMF staff estimates.Notes: Robust standard errors in parentheses. * denotes significance at 10% level, ** at 5% level, and *** at 1% level. All models include time dummies and outliers are excluded.
Table 5.

Determinants of Inflation

(nonlinear model estimated by nonlinear least squares)

article image
Source: IMF staff estimates.Notes: Robust standard errors in parentheses. * denotes significance at 10% level, ** at 5% level, and *** at 1% level. All models include country fixed effects and time dummies and outliers are excluded.

15. The standard determinants of inflation are confirmed, but there is little evidence of a price convergence effect. Inflation is positively correlated with its lag, pointing to persistence. Inflation increases with the output gap and decreases when the nominal effective exchange rate appreciates. However, the initial relative price level has no significant impact on inflation (in line with the more recent literature).

16. Less efficient labor market institutions increase the persistence of inflation. Inflation is more persistent when employment protection is high, collective bargaining is characterized by intermediate coordination, and union density is high. The presence of these interactions is presumably explained by the adjustment or indexation of wages to lagged inflation. High employment protection, high union density and some coordination of unions in collective bargaining give workers more market power to negotiate increases in their wages that compensate for high past inflation or high expected inflation (to the extent that future inflation is predicted based on past inflation). The relationship with the coordination in bargaining is non-linear, in the sense that both low and high coordination would lead to less inflation persistence than intermediate coordination. In the case of low coordination, workers have little market power, while in the case of very high coordination, the unions which recognize their market power take into account the effect of their wage demands on inflation and unemployment (the argument of Calmfors and Driffill, 1988). This high persistence that results from inefficient labor market institutions will lead to excessively high inflation when shocks to inflation are positive, such as during the boom in Spain or when oil prices increase. In contrast, in cases of negative shocks to inflation, inflation may remain excessively low for a period of time.

17. Collective bargaining systems with intermediate coordination are also less suited to face supply shocks (such as oil price shocks). Oil and raw materials price shocks are more likely to be accommodated by wage increases when the degree of coordination in collective bargaining is intermediate. The argument is similar to the one made above.

18. There are a few noteworthy non-results as well. First, product market regulation does not seem to have a robust impact on inflation dynamics.16 Theoretically, the effect of product market regulation could be ambiguous (Aghion, 2002). Indeed, when competition is low, firms faced with an increase in their costs could decide to either use their market power to raise prices (and thereby protect their profit margins) or to absorb the shock by reducing their profit margins to maintain market share. When competition is high, however, profit margins are very low and firms are more likely to be forced to raise prices when faced with increases in costs. In practice, the PMR indicator used in the analysis may also be an imperfect measure of market power. Second, none of the interactions for the output gap turned out to have a robust impact on inflation.17 One issue here might be the measurement of the output gap. The output gap is unobservable, notoriously difficult to measure, and its estimates are subject to frequent posterior revisions. Although the data are taken from a common source (IMF World Economic Outlook), methodologies used to estimate the output gaps may differ across countries.

19. These results are quite robust. The results are robust to different estimation methods, excluding outliers, dropping one country at a time, and different model specifications (linear or non-linear). A very stringent test, which tests all permutations of variables and calculates the robustness of variables by the frequency with which they appear in the best-fitting models, was implemented (through a publicly available program called “R”). Of all the variables, the most robust are the lagged inflation, the change in the nominal effective exchange rate, the output gap and the interaction of the lagged inflation with employment protection. The interaction of lagged inflation with intermediate coordination in collective bargaining and union density also appear in a good fraction of the best fitting-models, though they are not as robust as employment protection. Other variables were not robust. Including future inflation in the model and instrumenting it with lags of inflation and the output gap (as is typically done in estimations of hybrid New Keynesian Phillips curves) confirms that inefficient labor market institutions (in particular high employment protection) increase inflation persistence. The channel is through increasing the impact of lagged inflation on expected future inflation (in first-stage results). Finally, there is no evidence of residual slope heterogeneity on the macroeconomic variables for Spain after we control for their interactions with the structural variables, suggesting that the responsiveness of inflation to its determinants is fully accounted for by differences in the labor and product market variables (Table 6).18

Table 6.

Determinants of Inflation: Additional Robustness Tests for Linear Model

article image
article image
Source: IMF staff estimates.Notes: Robust standard errors in parentheses. * denotes significance at 10% level, ** at 5% level, and *** at 1% level. All equations are estimated using instrumental variables, and including country fixed effects and time dummies.

20. The magnitude of the impact of institutions on inflation and inflation differentials was relatively large. Over 1999-2009, the inflation differential of Spain relative to the 10 euro area countries that are in the estimation sample was on average 0.7 percentage points. Of this, 0.3 to 0.5 percentage points can be attributed to the specifics of the Spanish labor market, depending on the model (Figure 7). The contributions of employment protection and intermediate bargaining system were respectively 0.2 and 0.30.5 percentage points per year, partially offset by the low union density. These are likely underestimates though since they are a simple average of first-year effects on inflation, and do not take into account the dynamic effects of lower inflation on subsequent years. A simulation of what Spanish inflation would have been, had Spain exhibited the best possible combination of labor institutions, suggests that inflation would have hovered somewhat below 2 percent, more than 1 percentage point below the observed level.19 Thus, as 1999-2008 was a period of boom and shocks to inflation were mostly positive, labor market institutions overall contributed to increase the inflation differential through increased persistence and amplification of supply shocks. While labor market institutions kept pushing inflation up in 2009 despite the crisis, their contribution to the inflation differential became neutral in 2010, reflecting the small lagged inflation differential in 2009. It is noteworthy however that the model does not explain very well the behavior of inflation in crisis years (e.g. in 2001 and 2009).

Figure 7.
Figure 7.

Spain: Impact of Labor Market Institutions on Inflation, 1999-2010

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A006

Source: IMF staff estimates using estimated linear model.1/ Other includes current and lagged changes in nominal effective exchange rate, country fixed effect, and residual.

D. Policy Implications

21. These results suggest the accumulated inflation differentials during the boom years in Spain were larger and more persistent than desirable, in large part due to inefficient labor market institutions. The impact of common (and probably asymmetric) shocks on inflation was amplified by the structure of collective bargaining which featured an intermediate level of coordination. Moreover, the shocks to inflation were made more persistent by less efficient labor market institutions than in the average of other euro area countries. Inflation persistence, as estimated by our model, is particularly high in Spain compared with other euro area countries (Table 7).

Table 7.

Marginal Effect of a Change in Regressor on Annual Inflation over 2000-2007 1/

article image
Source: IMF staff estimates.

Includes direct and indirect effects.

22. These results suggest that the reforms, such as those recently, of collective bargaining and employment protection could improve inflation performance in Spain. The 2010 reform eased dismissal costs and criteria, and granted firms greater flexibility to opt out of collective agreements. In June 2011, collective bargaining was further reformed toward greater firm-level flexibility, through establishing the prevalence of firm-level agreements, especially over provincial ones; reducing the possibility of indefinite extension of previous agreements when social partners cannot agree on a new agreement; and further easing opt-outs of collective agreements. These reforms were first and foremost needed to improve labor market performance, as the persistent high wage growth during the Great Recession has put the burden of adjustment to the crisis on employment, pushing the unemployment rate to above 20 percent. But they could also contribute to reduce the persistence of inflation in Spain and help Spain achieve lower price growth relative to its partners to restore competitiveness without excessive contraction of absorption. If they fail to achieve these results, they will need to be strengthened.

References

  • Aghion, P., N. Bloom, R. Blundell, R. Griffith and P. Howitt, 2002, “Competition and Innovation: An Inverted U Relationship,” NBER Working Paper No. 9269 (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Andersson, M., K. Masuch, and M. Schiffbauer, 2009, “Determinants of Inflation and Price Level Differentials Across the Euro Area Countries,” ECB Working Paper Series No. 1129.

    • Search Google Scholar
    • Export Citation
  • Angeloni, I. and M. Ehrmann, 2007, “Euro Area Inflation Differentials,” The B.E. Journal of Macroeconomics, Vol. 7, Issue 1, Article 24.

    • Search Google Scholar
    • Export Citation
  • Bassanini, A. and R. Duval, 2009. “Unemployment, institutions, and reform complementarities: re-assessing the aggregate evidence for OECD countries,” Oxford Review of Economic Policy, Oxford University Press, vol. 25(1), pages 4059, Spring.

    • Search Google Scholar
    • Export Citation
  • Beck, G., K. Hubrich and M. Marcellino, 2009, “Regional Inflation Dynamics Within and Across Euro Area Countries and a Comparison with the United States,” Economic Policy, pp.141184.

    • Search Google Scholar
    • Export Citation
  • Biroli, P., G. Mourre, and A. Turrini, 2010, “Adjustment in the Euro Area and Regulation of Product and Labour Markets: An Empirical Assessment,” CEPR Discussion Paper Series, No. 8010.

    • Search Google Scholar
    • Export Citation
  • Bowdler, C. and L. Nunziata, 2007. “Inflation Adjustment and Labour Market Structures: Evidence from a Multi-country Study,” Scandinavian Journal of Economics, Wiley Blackwell, vol. 109(3), pages 619642, 09.

    • Search Google Scholar
    • Export Citation
  • Calmfors, L. and J. Driffill, 1988, “Bargaining Structure, Corporatism and Macroeconomic Performance,” Economic Policy, Vol. 3, No. 6, pp.1361.

    • Search Google Scholar
    • Export Citation
  • Correa-López, M., A. Garcia-Serrador and C. Mingorance-Arnaiz, 2010, “Product Market Competition and Inflation Dynamics: Evidence from a Panel of OECD Countries,” BBVA Research Working Papers, Number 10/26.

    • Search Google Scholar
    • Export Citation
  • de Haan, J., 2010, “Inflation Differentials in the Euro Area: A Survey,” in The European Central Bank at Ten, by J. de Haan and H. Berger (eds), Springer.

    • Search Google Scholar
    • Export Citation
  • ECB, 2005, “Monetary Policy and Inflation Differentials in a Heterogeneous Currency Area,” Monthly Bulletin, May 2005, pp. 6177.

  • Honohan, P. and P. Lane, 2003, “Divergent Inflation Rates in EMU,” Economic Policy, pp. 357394.

  • Honohan, P. and P. Lane, 2004, “Exchange Rates and Inflation under EMU: An Update,” Economic Policy Web Essay. Available at http://www.economic-policy.org/responses.asp.

    • Search Google Scholar
    • Export Citation
  • Rabanal, P., 2009, “Inflation Differentials Between Spain and the EMU: A DSGE Perspective,” Journal of Money, Credit and Banking, Vol. 41, No. 6, pp. 11411166.

    • Search Google Scholar
    • Export Citation
1

Based on joint work by Florence Jaumotte and Hanan Morsy.

2

Cyprus, Malta, Slovakia and Slovenia are excluded from the comparison since they joined the euro area much later (in or after the mid-2000s).

3

The exception is Luxembourg where the core inflation differential was small.

4

One exception is primary activities which has strong ULC growth and return on capital relative to the Spanish average but not a large differential on the value-added deflator relative to the euro area.

5

The rate of return on capital is only one component of UGOS. UGOS is the product of the rate of return on capital and of the capital to output ratio (i.e. the inverse of capital productivity).

6

This review of the literature draws on de Haan (2010).

7

There is more evidence that this is a relevant factor for new EU member states.

8

Indeed, the higher inflation leads to two opposite impacts on the output gap. On the one hand, in the context of the EMU, where nominal interest rates have converged, the higher inflation tended to reduce real interest rates and hence push the output gap even higher, sustaining upward pressure on prices in countries which already had high inflation. On the other hand, the higher inflation leads to a real exchange rate appreciation which gradually reduces export demand, thereby reducing the upward pressure on price developments.

9

Another related source of undesirable inflation differentials is rigidities affecting the price and wage formation mechanisms (e.g. low frequency of price adjustments).

10

Inflation and the output gap are from the IMF World Economic Outlook, nominal effective exchange rates from the IMF International Financial Statistics and the relative price level from Eurostat and the Penn World Tables. The indicator of coordination in collective bargaining takes the value 1 for uncoordinated systems, 2 for systems with intermediate coordination, and 3 for highly coordinated systems, and is from Bassanini and Duval (2009). Union density is the OECD measure of the share of workers affiliated to a trade union. Employment protection is measured by the OECD summary indicator of employment protection legislation. Product market regulation is measured by the OECD summary indicator of regulatory impediments to product market competition in seven non-manufacturing industries, including gas, electricity, post, telecoms, passenger air transport, railways, and road freight.

11

Spain also fares much less well on some subcomponents of the indicator, such as administrative burdens on start-ups and the regulation of professional services.

12

A more refined way to measure external shocks would be to control for imported good prices, distinguishing between oil and non-oil, and to interact these respectively with the share of oil-refined products in the consumption basket and with the degree of openness of the country. An element missing in the analysis is the role of indirect taxation and government-set prices in explaining inflation developments.

13

Correa-López et al. (2010) follow a similar approach but with a smaller set of structural variables. Biroli et al. (2010) also investigate the impact of labor and product market variables on the inflation dynamics. However, they only test the impact of one variable at a time and do not test all channels simultaneously.

14

The sample countries are Austria, Belgium, Finland, France, Germany, Ireland, Italy, Netherlands, Portugal, and Spain. Greece and Luxembourg could not be included because the collective bargaining indicator is not available for them.

15

The lagged inflation, the output gap, and all their interactions with the labor and product market variables are instrumented. The set of instruments includes the second lag of inflation, the first and second lags of the output gap, and the first lag of all interaction terms. The tests of whether instruments are strong and valid are satisfied in all cases.

16

Several studies find some impact of product market regulation on the inflation dynamics, e.g. Biroli et al. (2010), Correa-López (2010) and Andersson et al. (2010). One difference between our study and theirs is the larger set of structural variables that are introduced simultaneously in our model. For instance, when employment protection is not included (like in Correa-López, 2010), there is some evidence that product market regulation increases the persistence of inflation. However, when both variables are present, employment protection dominates product market regulation. Using the more comprehensive OECD product market variable (only available for a shorter time period) and disaggregating it into its different sub-indicators did not show a significant impact either.

17

While some interactions between structural variables and the output gap were marginally significant in some specifications, they usually became insignificant once the model was reduced in an iterative process to keep only significant variables. Allowing different coefficients for positive and negative output gaps did not improve the robustness of the results.

18

Other robustness tests were conclusive, including: (i) allowing a downward trend in the impact of the output gap on inflation, reflecting the fact that the credibility of monetary policy has improved over time resulting in an anchoring of inflation expectations and a reduced sensitivity of inflation to the output gap, beyond the effect of institutional changes; (ii) testing whether the levels and changes of the structural variables affect inflation beyond their effects through interactions with macroeconomic variables (they typically do not); and (iii) re-estimating the model including the Great Recession period (2008-2009) and assuming that structural variables for which data are not available over this period remained broadly constant (see Table 6).

19

This simulation is based on the linear model and takes into account the impact on inflation in subsequent years through lower lagged inflation but not through lower output gap (since we do not have a model of the output gap).

Spain: Selected Issues
Author: International Monetary Fund
  • View in gallery

    Spain and Euro Area: Headline Inflation

    (year-on-year percent change)

  • View in gallery

    Spain: Contributions of Inflation Components to Inflation Differential with Euro Area

  • View in gallery

    Spain: Accumulated Price Differences Relative to Euro Area

    (index, 1999=100)

  • View in gallery

    Spain: Inflation Differential with Euro Area by Value Added Sector

  • View in gallery

    Selected Euro Area Countries: Accumulated Price Differences Relative to Euro Area, 1999-2010

    (index)

  • View in gallery

    Spain and Euro Area: Determinants of Inflation

  • View in gallery

    Spain: Impact of Labor Market Institutions on Inflation, 1999-2010