Journal Issue
Share
Article

Luxembourg: Selected Issues

Author(s):
International Monetary Fund. European Dept.
Published Date:
May 2014
Share
  • ShareShare
Show Summary Details

EXTERNAL DEVELOPMENTS, COMPETITIVENESS, AND LABOR MARKET POLICIES1

Luxembourg maintains a comfortable external position, supported by strong export services underpinning cross-border financial activities. Yet, price and cost developments have been far less benign. If sustained in the future, those developments could hamper Luxembourg’s ability to diversify its economy, at a time when the financial landscape is undergoing significant changes. In order to turn those trends around, some features of the labor market, in particular the automatic wage indexation mechanism, will need adjustment.

A. Background and Overview

External developments

1. Luxembourg has been relatively resilient to the euro area crisis as a financial center. As the world's second largest investment fund center and a leading place for private banking and reinsurance business, Luxembourg has maintained its unique competitive advantage in the euro area and continued to explore new businesses. Since 2011, the financial sector (service exports) has provided a solid buffer to growth, when domestic demand and other exports were more depressed. Moreover, the Luxembourg authorities have developed a diversification strategy around several areas, including ICT. As a result, some internet start-ups and data centers have settled in Luxembourg, which has also supported employment and exports.2 Over the past decade, sound economic policies, including a prudent fiscal stance, have buttressed the economy and allowed Luxembourg to maintain a generous social safety net.

Contribution to GDP Growth

(In percent)

Sources: Haver Analytics, IMF Staff calculations.

Current Account by Components

(In percent of GDP)

Sources: Haver Analytics, IMF Staff calculations.

2. Subdued demand from the euro area led to a moderate correction of the current account surplus in the past two years. The current account balance fell to 5.8 percent of GDP in 2012 from its 2010 peak of 7.7 percent, on the back of weak good exports to the rest of euro area. In 2013, the current account is estimated to have moderated further to 5.2 percent of GDP, as a result of less buoyant service exports. However, there are mitigating factors: service exports still contribute half of the total surplus, and good exports have picked up recently. Income outflows, mainly salary remittances of cross-border workers and investment income, remained strong and are close to the pre-crisis peak.

Exports and Imports Growth since 2011

(In yoy percent change)

Sources: Haver Analytics, IMF Staff calculations.

3. Mirroring sustained current account surpluses, the International Investment Position (IIPs) has continued to build up. Luxembourg’s net IIP position reached about 180 percent of GDP by end-2013, one of the highest in the euro area, and to a large extent reflecting activities associated with a financial center.3 Despite net portfolio inflows from investment fund activities, the overall financial account balance vis-à-vis the rest of world has registered a deficit largely due to “other investment” activities. A new survey methodology was implemented from 2011Q4 onward, resulting in an overall upward revision to the net FDI position and statistical changes in the composition of the IIP, due to an upward revision to the FDI gross asset position and a downward revision to the gross liability position.

International Investment Position

(In percent of GDP)

Sources: Haver Analytics, IMF Staff calculations.

Market structure and shares

4. Luxembourg registers a significant trade deficit in goods. More than half of Luxembourg’s good exports (steel and processed food) go to neighboring countries in advanced Europe, and over 70 percent of imports (lumpy investment and consumption goods) are from the euro area. As a small open economy with a modest production base, the traded good balance has been registering a deficit of around 10-13 percent of GDP for the past decade. Trade with other euro area members has been increasing, against the overall trend of declining intra-EA trade since the mid 2000s. The top three trading partners are Luxembourg’s direct neighbors, Germany, Belgium, and France. But the share of China has been increasing; it is now the third largest trading partner on the import side.

Luxembourg's Top 10 Exports Destination: 2007-12

Share of total exports (goods)

Sources: UNCTAD and IMF Staff calculations.

Luxembourg's Top 10 Imports Destination: 2007-12

Share of total Imports (goods)

5. But Luxembourg also maintains a sizeable market share in service exports, driven by financial sector activities, which are at the root of its significant current account surplus. Over the past decade, financial services have been Luxembourg’s largest exporting sector, and the country accounts for about 15 percent of global financial exports. Besides a large investment fund industry, the banking sector operates diversified business lines on a range of cross-border activities.4 While the sector faces substantial challenges, the authorities are turning their diversification efforts toward non-euro area markets, especially emerging markets. A specific recent emphasis has been on Renminbi business, with several Chinese banks setting up in Luxembourg to benefit from access to the single market. Other sectors, such as IT and communication, have been fast growing, but market shares remain small.

World market shares of service exports

Average 2000-2011, in percent

Sources: UNCTAD, IMF Staff calculations.

Market Share of Financial Service Exports: 2011

B. External Competitiveness

6. Despite a healthy current account position, price developments have been less benign. Luxembourg’s real effective exchange rate (REER) appreciated markedly compared to other euro area peers since the beginning of the crisis. Measures based on Unit Labor Cost (ULC) and GDP deflators have increased by 8 and 6 percent since 2008, while they have declined by 0-5 percent in Germany, France, and Belgium during the same period. If this trend were to continue, it could eventually have a negative impact on external competitiveness from higher labor and living costs. In addition, the CPI-based REER has been showing signs of decoupling from the ULC- and GDP-based measures recently with less appreciation, suggesting that imported inflation from trading partners dropped significantly.5

ULC-based REERs for Total Economy

(In index number, 1999Q1 = 100)

Sources: Haver Analytics, IMF Staff calculations.

Luxembourg: Real Effective Exchange Rates

(In index number, 1999Q1 = 100)

7. The progression in Luxembourg ULCs has outpaced that of most euro area members. Total economy ULC in Luxembourg rose by 11 percent since end-2009, and more than 30 percent since the beginning of the global crisis.1 Labor hoarding, in particular in the financial and manufacturing sectors at the onset of the crisis, led to a sharp decline in labor productivity, but wage inflation has also been particularly dynamic (see next section). As a consequence, wage growth, partly resulting from the automatic wage indexation, has consistently outstripped productivity gains for a number of years. More recently, labor cost increases have been moderated by a less dynamic labor market as well as by the temporary adjustment of the automatic wage indexation mechanism, which puts a cap on the frequency of wage adjustments until end-2014 (see next section).2

Total Economy ULC Developments: 2000 = 100

Sources: Statec and IMF staff calculations.

Luxembourg: Changes in ULC

8. The pattern of ULC developments is also confirmed at the sectoral level. It is well known that for exports to grow, production cost in tradable sectors must be competitive relative to trading partners, other things equal. In the case of Luxembourg, manufacturing and financial services are the two largest tradable sectors, and they both experienced large ULC increases in past years. The annual ULC change in both sectors has been close to 4 percentage points higher than the EA average over 2010-2012. If these trends are not reversed, there is a risk of pricing out core activities with the largest share of employment, in particular in comparison with other financial centers in the EU like the UK.

Annual ULC Change: Manufacturing

(In percent, relative to EA17, 2010-2012)

1/ Industry ULC for Belgium.

Sources: Eurostat, STATEC, and IMF.

Annual ULC Change: Financial and Insurance

(In percent, relative to EA17, 2010-2012)

C. The Role of Labor Market Policies

Implications for competitiveness

9. The recent crisis has shown how labor market policies can have important implications on competitiveness and macroeconomic stability for euro area members. Up to the crisis, many euro area periphery economies experienced a large deterioration in competitiveness due to rising labor costs. Wage growth surpassed the rise in labor productivity, and was accompanied by consumption booms, which led resources to be allocated to non-tradable sectors and eroded competitiveness. Cross-country experiences suggest that some labor market features, including in relation to wage determination and some social benefit schemes, can contribute to the deterioration of competitiveness.

10. While Luxembourg enjoys a comfortable external position, wage indexation plays a large role in price developments and ultimately competitiveness. This type of wage adjustment mechanism exists only in a few other countries in the euro area, namely Belgium, Cyprus, and Malta. The principle is to automatically adjust wages according to an index that tracks cost of living standards, calculated in Luxembourg using the CPI. Once the cumulative increase in the CPI index reaches 2.5 percent, all wages are adjusted. Up to 2011, the adjustment could happen up to twice a year. The traditional indexation mechanism was suspended in 2012, and a temporary regime is in place until December 2014. During this period, the initial indexation (which should have taken place in March 2012) was postponed to October 2012, and a minimum period of twelve months must elapse between two indexations. At the current juncture, the next indexation can not happen before October 2014.

11. The automatic indexation of wages may be partially responsible for the strong dynamism of labor costs and inflation in recent years. Strong wage increases can trigger higher inflation than in euro area partners, as the rise in inflation automatically causes wages to increase, and as a second round response, the subsequent rise in wages increases inflation further—a process that can result in labor cost increases that diverge from productivity gains when labor productivity declines, as has been the case since the crisis in Luxembourg.1 Moreover, the annual wage inflation has been relatively higher than elsewhere in the euro area, where adjustments are ongoing.

Annual Wage Inflation: Total Economy

(In percent, relative to EA17)

Sources: Eurostat and IMF.

Annual Wage Inflation: Financial and Insurance

(In percent, relative to EA17)

Sources: Eurostat and IMF.

12. The inflation differential between Luxembourg and the euro area has been relatively persistent, at close to 1 percentage point on annual basis. This wedge has been positive since the inception of euro, even during recessions. The bi-annual review of the minimum wage has also contributed, with the minimum wage being one of the highest among advanced economies.

Core inflation: Luxembourg and Euro Area

(In percent, qoq annual moving average)

Sources: Eurostat and IMF.

Official Minimum Wage, 2013

(In euros per month)

Sources: Eurostat

Note: After 2.5 percent increase on Oct 1 2013 in Luxembourg.

13. An econometric analysis suggests that at least half of the inflation differential with the euro area can be attributed to the automatic indexation mechanism. In a model linking Luxembourg inflation to the euro area inflation and idiosyncratic effects such as the automatic indexation and the review of minimum wages, the latter effects are found to contribute to 1 percentage point to the annual inflation rate.2 In this context, the temporary arrangements in place until end 2014 limiting the frequency of adjustment are welcome, but a permanent system less conducive to high inflation persistence should be designed to preserve competitiveness and promote a sound macroeconomic environment. This is especially important in a context where, unlike in neighboring countries, the increase in labor compensation has been accompanied by a decline in labor productivity since the crisis.

Luxembourg inflationDiff 1/LevelLevel
Constant0.55***1.67***1.25***
Euro area inflation0.45***0.58***
Output gap0.020.03
Indexation (-1)0.53***0.59***
Minwage (-1)0.370.40**
Observations474747
R20.130.280.48
AR(1 ) error coef.-0.36***-0.52***
Durbin-Watson2.611.972.09
Sources: IMF staff estimations.

Differential with euro area inflation

”***” <i percent, “**” < 5 percent, “*”< 10 percent.

Sources: IMF staff estimations.

Differential with euro area inflation

”***” <i percent, “**” < 5 percent, “*”< 10 percent.

Compensation and real value added per worker

Policy agenda

14. Alternative growth drivers should continue to be explored, along with continuing social and economic reforms. Besides a prudent fiscal stance and a low-tax business-friendly environment, Luxembourg’s success has also reflected its ability to react quickly to changes in the global environment, often giving it a first-mover advantage, in particular in the financial sector. The new government has stated its commitment to sustain the long-term development of Luxembourg as a financial center in general and to continue to pursue diversification beyond the financial sector. However, rapidly increasing labor costs—themselves reflecting specific features of the labor market—could have significant repercussions on the ability of the economy to react, in particular as the financial sector adjusts to significant EU-driven changes. Delaying reforms in these areas could therefore not only erode competitiveness, but also hamper the growth potential in the long run.

15. The trend increase in labor costs should be halted to preserve competitiveness, and in particular to avoid placing barriers to diversification beyond the financial sector. Luxembourg’s efforts to limit the negative side-effects of indexation—in particular by limiting the adjustment to 2½ percent a year—have helped contain rising unit labor costs. While wage indexation is seen as an intrinsic part of the social contract, several options could be explored to help reduce its undesirable consequences on competitiveness:

  • Making the limit of one indexation round per year a permanent feature of the mechanism after the current arrangement expires, but setting the annual cap closer to the ECB price stability objective (close but below 2 percent) than the current 2½ percent threshold. This would not preclude additional wage increases negotiated between employees and employers, but would allow for wages to reflect more closely labor productivity developments.

  • Supplementing the current mechanism with escape clauses when inflation among the main trade partners falls significantly below the annual cap to automatic indexation.

  • Modifying the reference index to exclude volatile prices (notably food and fuel prices).

16. Measures to strengthen labor skills and the business environment will further support Luxembourg’s diversification efforts. In particular, policies should aim at:

  • Continuing to reduce the gap between workers’ skills and private sector needs. Human capital accumulation can be further supported by life-long learning and continued attention to improving vocational and formal education, where Luxembourg ranks low compared to OECD peers.3 Efforts should continue to be made to match workers’ skills with private sector needs, including through support for training in the areas of targeted growth, in close collaboration with the private sector. A review of the current education system should complement these efforts, with a view to reducing dropout rates and increasing efficiency.4

  • Reduce barriers to entry and competition. Luxembourg also has relatively more stringent product market regulations (PMR) compared to other EU members, notably on professional services and network industries (e.g., retail distribution).5 Further relaxation of these regulations would support resources to move freely across sectors, especially those sectors that will be Luxembourg’s future sources of growth (text table). The emergence of new firms could be facilitated by improving administrative processes to start a business.

Summary of Recommendations for Luxembourg: 2014 OECD Going for Growth Interim Report
Areas of improvementRecommendations
Improve the design and integration of unemployment benefits and active labor market policiesAddress early exit of older workers from the labor market
Reduce impediments to job creationReview some provisions of employment protection legislation
Reduce barriers to entry and competitionReduce entry barriers to professional services and network industries
Sources: OECD.
Sources: OECD.

17. Overall, deepening labor and product market reforms are imperative to enhance sustainable growth and boost competition, which requires addressing long-term unemployment, fostering productivity growth, and more efficient resource allocation. The economic downturn has brought to the core the burden of these policies on the fiscal purse, as well as their unintended disincentives to work and impediments to competition, which remain to be addressed.

Appendix: Identifying the Potential Impact of Wage Indexation and Minimum Wage Review on Luxembourg’s Inflation

A. Background and model specifications

1. Inflation in Luxembourg has been higher than many of its euro area peers. In general, as a small open economy subject to relatively large external shocks, it is not unusual for Luxembourg to exhibit more economic volatility. However, the inflation differential between Luxembourg and the euro area average has been quite persistent over time, at close to 1 percentage point on annual basis since the inception of euro. This raises the question as to whether there are any structural or country-specific factors other than cyclical conditions that may contribute to this fact.

2. In this context, the analysis explores the potential impact of the automatic wage indexation mechanism and minimum wage review, two distinct features of the Luxembourg labor market, on its inflation dynamics. In general terms, persistence of inflation (or inflation differentia) may be defined as the tendency of inflation to revert slowly to its equilibrium or long run level after a shock. In Luxembourg, all wages are adjusted according to an index that tracks cost of living standards, which is calculated using the CPI. Once the cumulative increase in the CPI index reaches 2.5 percent, all wages are adjusted and it can happen up to twice a year.1 In addition, minimum wages are reviewed at the beginning of every two years. Both can be considered as exogenous shocks to domestic inflation.

3. To help explain the inflation differential with the euro area and inflation dynamics, two sets of time series model are used to estimate inflation differential and inflation in levels. To start with, we consider a simple model as follows:

where “inflation differential” represents the difference between Luxembourg inflation and the euro area inflation (qoq annualized), and “Indexation” and “Minwage” are time dummies indicating the quarter when wage is adjusted. In addition, the constant captures the part of the wedge between Luxembourg and euro area inflation that is not related to those two variables.

4. An alternative model is also estimated using a modified Philips Curve. As euro area inflation could be an importance source of imported inflation to Luxembourg, which is suggested by the Granger causality test (see appendix table), it is included in the estimation equation together with other explanatory variables as below.

5. Inflation in Luxembourg and euro area are measured by HICP excluding energy and unprocessed food. Both equations are estimated using quarterly data from 2001 to 2013.

Granger Causality Test
Null Hypothesis:F-Statistic Prob.
EA inflation does not Granger Cause LUX inflation5.860.01
LUX inflation does not Granger Cause EA inflation1.440.25

B. Results summary

6. The estimation results are presented in the table below. First, the inflation differential is well captured by both the constant and indexation, but to a less extent by the minimum wage review dummy (but close to 10 percent significance level). Results suggest that indexation alone could increase explain ½ percentage point of the inflation differential. Second, spikes in inflation are found to happen the quarter directly following an episode of automatic wage indexation, suggesting indexation have a significant impact on the inflation gap with the euro area.

Luxembourg inflationDiff 1/LevelLevel
Constant0.55***1.67***1.25***
Euro area inflation0.45***0.58***
Output gap0.020.03
Indexation (-1)0.53***0.59***
Minwage (-1)0.370.40**
Observations474747
R20.130.280.48
AR(1) error coef.-0.36***-0.52***
Durbin-Watson2.611.972.09
Sources: IMF staff estimations.

Differential with euro area inflation

”***” <1 percent, “**” < 5 percent, “*”< 10 percent.

Sources: IMF staff estimations.

Differential with euro area inflation

”***” <1 percent, “**” < 5 percent, “*”< 10 percent.

Inflation Differential and Wage indexation

QoQ annualized inflation, in percent

7. Finally, inflation dynamics in Luxembourg are better captured by the alternative models with increased fitness and reduced serial correlation (see table). As suggested by the Granger Causality test, inflation from the euro area truly contributed to the domestic inflation with a 50 percent passthrough, e.g., one percentage increase in euro area inflation would bring up inflation in Luxembourg by around ½ percentage point. In addition, while the sign of output gap is correct, the impact is small and insignificant. More importantly, there is evidence showing both indexation and minimum wage review have a substantial impact on inflation. If indexation and minimum wage review happen within the same year, they can contribute up to one percentage point of inflation differential with the euro area, other things equal.

References

Prepared by Shengzu Wang (EUR).

Unlike other parts of the euro area, employment growth has remained positive in Luxembourg, reaching more than 2 percent in both 2011 and 2012.

These include banking and investment fund activities, as well as activities of Special Purpose Entities (SPEs). SPEs include for example in-house banks for corporate groups, or proceeds from international issuance of securities in Luxembourg by multinational corporations.

See also the Selected Issues Paper: “The Financial Sector: Strengths and Challenges”.

The CPI basket, for example, has a much higher content in imported prices than the GDP deflator.

ULCs of Luxembourg’s major trading partners like France, Germany, and Belgium increased by 5-10 percent since end-2009, and 10-20 percent since the global financial crisis.

Under a temporary regulation in place until December 2014, automatic wage indexation is limited to once every 12 months—as opposed to up to twice a year in the main framework.

Lünnemann, Patrick and Wintr, Ladislav, (2010), “Downward wage rigidity and automatic wage indexation: Evidence from monthly micro wage data”, No 48, BCL working papers, Central Bank of Luxembourg.

See technical appendix for a detailed description of the empirical analysis and model specifications.

Luxembourg’s education system faces two unique challenges. It has to provide students with the skills that are in demand in a labor market with strongest employment growth in the financial sector and related services; and equip students with the language skills necessary to communicate effectively and work within a country with three official languages (OECD Working Paper 778).

“Global Competitiveness Report 2013-14”, World Economic Forum, Geneva.

“2014 OECD Going for Growth Interim Report”, OECD.

Under a temporary regulation in place until December 2014, automatic wage indexation is limited to once every 12 months—as opposed to up to twice a year in the main framework.

Other Resources Citing This Publication