The Luxembourg Economy Kaleidoscope (2004)

The Luxembourg Economy

in 2003-2004

A kaleidoscope

The Luxembourg

Economy

in 2003-2004

A kaleidoscope

The Luxembourg Economy

in 2003-2004

A kaleidoscope

Summary

1. Background: a fewstructuralelements  9

1. Exceptionalgrowthfrom1985to2000 10
2. Job-intensivegrowth 13
3. From2001to2004:reversalintheeconomiccycleandrecovery 18
4. Labourproductivityandemployment:contrastingtrends 20
5. Useofforeignlabour:cross-borderworkersandimmigrants 29
6. Activityrates:delayedentryintothelabourmarket,earlyretirementandincreaseinthefemaleactivityrate 32
7. Structureofvalue-added:the"weight"offinancialservices 34
8. Businessdemography: a dynamicdevelopment 41
9. Externaleconomicrelations:importanceofforeigndirectinvestmentandcurrentaccountsurplus 44
10. Wagerestraint 48
11. Publicfinances: a fragilebalance? 52
12. Povertyrateandincomedistribution:relativestability 63

2. Theeconomicsituationin2004  71

Impressum

Publishing manager: Serge ALLEGREZZA

Editor: Paul ZAHLEN

Contributions: Ferdy ADAM, Martine HILDGEN, Jean LANGERS, Bastien LARUE,Véronique SINNER, Guy SCHULLER

Technical coordination: Guy ZACHARIAS

Design and layout: INTERPUB' Printing: IMPRIMERIE CENTRALE Translation: EUROSCRIPT

This publication has been produced with the support

of the Ministry of Foreign Affairs and Immigration – Luxembourg Presidency of the Council of the European Union

1. Theinternationaleconomicsituation:Europelaggingbehind 72
   1. UnitedStates:growthanddeficits 72
   2. Asia:ChinaandJapan:increasinglycloseties 73
   3. Europe:stillbringinguptherear 74
2. EconomicsituationinLuxembourg 75

1. 2004: a satisfactoryyearexceptforthelabourmarket 75
2. Forecasts:sustained,butnotexceptionalgrowth 79
   3. Recentactivityin a numberofeconomicsectors:mixedperformance 82
      1. Industry:2004was a goodyear 82
      2. Construction:buildingandcivilengineeringmovinginoppositedirections 84
      3. Commerce:slowdown  87
      4. Financialsector:goodperformancebyUCIsandsustainedgrowthoftheinsurancesector 89
   4. Price sandwages:inflationrestrained,moderategrowthinwagecosts 95
   5. Externaleconomicrelations:balanceofservicesshowing a recordsurplus,growingcurrentaccountbalance 98
   6. Employmentandunemployment:employmentrecovering,butunemploymentratestillontheincrease 103

Central Service for Statistics  3 Thematic reports  109

and Economic Studies

3.1 Lisbon Strategy and structural indicators: Luxembourg's situation and margin for manuvre 110

13, rue Erasme 3.2. Competitiveness indicators 132

B.P. 304 3.3 Innovation and R&D: strengths and weaknesses of Luxembourg 141 L-2013 Luxembourg 3.4 Information and communication technologies (ICT) in 2003-2004: towards an information society 145

3.5 Stock market indices and economic growth 154

E-mail: info@statec.etat.lu 3.6 International business cycle indicators: their forecasting capability for the Luxembourg economy 156 Internet: www.statec.public.lu 3.7 GDP, value-added and FISIM: the case of Luxembourg 158

Tel.: (+352) 478-4219 3.8 Price s and price levels: international comparisons 162

3.9 Expenditure by cross-border workers in Luxembourg 171

April 2005 3.10 Demography: dynamic growth, questions of mobility and housing 173 ISBN 2-87988-059-9 3.11 Wages and salary costs: how are they measured? 187

3.12 Standard of living, poverty and social exclusion in international comparisons 201

Copies are authorised provided the source is mentioned.

Foreword

There are several reasons for launching a new publication on Luxembourg's economy and society. STATEC's three main motives are the:

• wish to communicate not only gross figures and databases, but also, and above all, analyses to make the observations intelligible;

• requirement to summarize the studies and research conducted by STATEC or by research institutes, sometimes jointly;

The Lisbon Strategy, descredited by its mediocre results,STATEC: statistical office

has become more realistic: it is no longer a question of beco-The Central Statistical and Economic Studies Office (STATEC) ming the most competitive economy in the world, but der vingis also changing. It should be borne in mind that STATEC has maximum benefit from science and technology to modernizea dual mission: the production of statistics on the one hand, the European social model. Each government is expected toand analysis and forecasting on the other.The first function, present a National Action Plan, in which it sets realistic targetsthat of a statistical office, has developed very strongly to com- for growth and employment and presents the policies that itply with the injunctions of Community regulation: the European proposes to implement. Sooner or later, it will be necessary toCommission and the Central Bank have developed a growing evaluate the effects of these policies and compare the targetsappetite for data, which they request in ever-greater volume announced with the results obtained: there too, harmonizedand at more frequent intervals.The production of quarterly

and reliable indicators are required. national accounts, a difficult and delicate exercise for a small

economy, is an illustration of this development.

• need to report regularly and comprehensively on the state of Luxembourg's economy and society and the outlook.

This year also offers a special and rare opportunity to launch the first book in this series: that of Luxembourg's Presidency

of the European Union, and especially the ratification of the Constitutional Treaty for the European Union, which will be

put to a referendum. STATEC wants to make its contribution

to the debate by setting the Grand Duchy, which has developed and prospered in a peaceful Europe, in a broader political

and economic context.

European economic governance

The Luxembourg presidency of the EU has carried out reforms of economic governance: the reform of the Stability and Growth Pact as well as the relaunch of the Lisbon Strategy.The exces- sive harshness of the common rules for managing national budgets was justified by the realization of the Economic and Monetary Union, and the introduction of the euro in a currency exchange area which was far from optimal.The reform decided by the European Summit in March 2005 will enable these rules to be applied more intelligently, in particular by taking account of the position of the economy in the business cycle and the nature of the structural reforms that public money is intended to promote. Harmonized budgetary statistics are the responsi- bility of statisticians who must demonstrate the utmost profes- sionalism and impeccable scientific impartiality in carrying out their work.

Lisbon-Luxembourg

Following on from the report "Une paille dans l'acier"  STATEC: research centre

("A straw in the steel") by Prof. L. Fontagné, STATEC was The second function, analysis and forecasting, is just as invited, together with the Competitiveness Observatory, important, because it enables the mass of data collected

to draw up the "Lisbon" National Action Plan, in particular to be understood, and made intelligible for use in determining by setting up a management dashboard of the Luxembourgeconomic policy. STATEC has improved its annual forecasting economy.This management dashboard, which uses data instruments, and is working on producing infra-annual macro- gathered or produced by STATEC, will take account of theeconometric models. Models for long-term forecasting of social, economic and environmental dimension. the country's population and economic potential have been

developed.These tools and results are decisive in explaining,

for instance, the financial sustainability of our welfare state, Break in the growth pattern, or minor turbulence? or land-use planning.Advances in econometrics and informa- The year 2004 probably marks the end of a period of an excep-tion technology are facilitating the analysis of rich individual tional, clearly perceptible slowdown in the Luxembourg busi-data on businesses and households. Many projects are ongoing ness cycle, which had interrupted an exceptional period of– often in partnership with research centres and universities – growth.Was this minor turbulence or a break in the growthin the field of labour and social cohesion as well as productivity pattern that we had previously experienced and to whichand new information and communication technologies or Luxembourg had become accustomed or had even taken entrepreneurship.

for granted? Although the result may be the envy of other

European countries, growth remained below its potential,This publication is intended for a wide audience, and is distri- too weak to stem the rising unemployment rate.The system buted as a book illustrating the way in which the economy is

of production, which is still insufficiently diversified, relies working in the short term, as well as its strengths and weak-

on financial activities, the results of which are tied to the nesses, the opportunities and hazards.The book attempts to fluctuations in share prices and the volatility of the financialavoid the esoteric jargon used by economists, the technical markets. terminology of statisticians, and the hollow language of tech-

nocrats, while making no concessions in terms of scientific rigour.

I hope you enjoy reading it, and invite your comments. Dr S.Allegrezza

Background:

a few structural elements

1.1 Graph 1.1.1

GDP per capita in PPS, 1960-2004 (EU-15 = 100)

Exceptional growth  211086000 EIUrUen-liat1en5dd States GNSweertihmtzeaernrlalyanndds from 1985 to 2000 111420000 LBJauepxlgaeinummbourg FPNrooarrntwucgaeyal

80

60 Source: European Commission (AMECO)

40 N.B.: PPS = Purchasing Power Standard

Graph 1.1.2

Differential of Luxembourg cycle compared to European cycle, 1960-2005

in %

10 GDP EU-15,

8 differential compared to trend

6 GDP Luxembourg,

4 differential compared to trend

2

0

-2

-4

-6

Source: European Commission (AMECO) -8

The reversal in the economic cycle in 2001 – GDP growth fact that during the 1960s – and to a lesser extent in the fact that the level of GDP per capita in France was slightlyIn addition, the GDP per capita includes categories of expendi- fell from 9% in 2000 to a little more than 1.5% in 2001 – 1970s – the growth of the Luxembourg economy does not higher than the EU average from 1960 to 2004 without tures, such as gross fixed capital formation, that only indirectly should not overshadow the fact that the average growth  seem to have been able to keep pace with the average of the sharp changes experienced in most other Europeancharacterize the economic well-being of households.To this of the Luxembourg economy from 1985 to 2004 was more the EU-15; this resulted in a fall in the level of GDP per capi- countries. is added that the general price level is a reflection not only of than 5% per year, exceeding by far the growth of the other ta in comparison with the EU average and helps explain both the price of the goods and services consumed by inhabitants countries of the EU-15, except for Ireland. the slight growth in real wages in Luxembourg in compari- One of the principal characteristics of the growth of the Luxem-of the country, but also includes capital goods, collective public

son with the EU-15 (see p. 192) and increased efforts made bourg economy is its volatility. In general, the economic cycleservices, etc.

The graph 1.1.1 indicating the change in the level of GDP  to diversify the Luxembourg economy in the 1960s. In addi- in Luxembourg follows that of the other European countries,

per capita in PPS from 1960 to 2004 (EU-15 = 100) provides tion, this phase of Luxembourg's economic history increases but with a greater range of variation in GDP (see graph 1.1.2The first bias resulting from the significant number of cross-

a synthetic picture of the long-term economic trend of the awareness of the fact that such a scenario could occur indicating the differential of the Luxembourg cycle comparedborder workers in the total workforce can be partially elimina- Luxembourg economy and of some industrialized countries.again;  with the European cycle).This is characteristic of a smallerted by using "gross national income"(GNI), which takes into Particularly noteworthy are the: economy open to outside influences, which makes it more account the flow of factors between Luxembourg and abroad,

• level of GDP of the EU-15 rising to that of the United States vulnerable to external shocks. that is, in particular, pay that is "exported" by cross-border

• quite exceptional nature of Luxembourg's growth  from 1960 to the mid-1970s and the decline in the level of workers into neighbouring countries. For example: in terms from the mid-1980s on; this growth reinforces a level  GDP per capita of the EU-15 in comparison with the United The GDP per capita is often used as an approximate measureof nominal value, GDP at current prices in Luxembourg was

of GDP that was already higher at the time than most  States beginning in the 1990s. Nevertheless, from a very of the average standard of living of a country, for internationalEUR 23,956 million in 2003, while the GNI was just EUR

of the other industrialized countries of Europe; long-term perspective this trend does not mean as much comparisons, among other things. It can be expressed in PPS21,206 million during that same year.

lost ground as is frequently attributed to it; (Purchasing Power Standard), which neutralizes differences in

• effects of the economic slowdown in Luxembourg from price levels. However, GDP per capita is only a very rough esti-But using GNI does not eliminate the second bias mentioned 1992 to 1996 and of the reversal in the economic cycle economic performance of Ireland, and to lesser extent, mate of the change in the standard of living. For instance, oneabove, that is, the fact that it includes, like GDP, categories of of 2001 (and less impressive growth from 2002-2004) that of Norway, beginning in the mid-1980s; of the reasons behind the higher level of the per capita GDPexpenditures, such as gross fixed capital formation, that only which are illustrated by the stagnation of the level of GDP index in Luxembourg is the significant number of cross-borderindirectly characterize the economic well-being of households. per capita of Luxembourg in comparison with the EU-15. effect of the reunification of Germany, which led to a sharp workers employed in the country. Cross-border workers repre-However, it must be remembered that these investments decline in the GDP of reunited Germany in comparison with sent almost 40% of salaried employment and about one-thirdconstitute a basis for future well-being. Still, individual effec- the EU. Subsequently, the GDP per capita continues to fall of all employment.These workers contribute to GDP and aretive consumption seems a more appropriate measure of slowly but steadily in comparison with the EU average, part of the labour market, but they are not included in the average economic well-being of households at a given time. which explains current concerns in German economic and population and consequently are not included in the denom -This consumption includes expenditures for household final political circles; nator of the GDP/population ratio. consumption expenditure as well as expenditures for individual consumption by public administrations, which is composed

of public services that benefit households individually, such as state-funded medical services.

Graph 1.1.3 1.2

GDP per capita and effective individual consumption per capita, 2002 (OECD = 100)

225000 Job-intensive growth 150

100

50 0

Effective individual consumption per capita

GDP per capita Source: OECD

Some countries, such as the United Kingdom and the UnitedThis explains why the index of GDP per capita in these coun- In Luxembourg, economic growth obviously had a positive a spectacular acceleration in the rate of job creation States, have an effective individual consumption index signifi-tries exceeds by far the OECD average (index of 205 for Luxem- impact on employment trends (see comparative graph 1.2.2 between 1997 and 2001 (a job growth rate of 5.6% in 2000 cantly higher than the GDP per capita (see graph 1.1.3). Bybourg and 129 for Ireland), but the differential between the on employment trends and GDP from 1985 to 2004). and again in 2001), parallel to exceptional GDP growth rate way of example, in the United Kingdom the GDP per cap ta individual consumption index and the OECD average is much Nevertheless, the intensive nature of employment in growth from 1997 to 2000 (GDP growth rate of 8%); this accelera- in 2002 was 13% above the OECD average, while per capitaless.This index is 159 (OECD = 100) in Luxembourg. In Ireland, in Luxembourg should be noted.The average annual rate  tion in the rate of job creation was accompanied by a very consumption was more than 27% above that average.Theseit even falls below the OECD average, and was at 96 in 2002. of job growth from 1985 to 2004 in Luxembourg was 3.2%,slight reduction in the unemployment rate;

figures reflect a structure of spending characterized by a subs- compared with under 1% in the EU-15.The number of jobs

tantially lower than average volume of investment per capitaIn summary, the differential in the standard of living of Luxem- in the Luxembourg economy rose from 161,000 in 1985 (of a deceleration in job creation beginning in 2002 with a and substantially higher than average individual consumption.bourg in comparison with other EU and OECD countries as sug- which 147,200 were salaried positions) to 293,500 (of whichgap of approximately one year compared to the economic

gested by GDP per capita is reduced to more equitable propor- 273,600 were salaried positions) in 2003. In December 2004,turnaround of 2001 and which was accompanied by an

In other countries, such as the Netherlands, Denmark, Ireland,tions if reference is made to more appropriate measures, such total domestic employment numbered 303,546 people. increase in the unemployment rate;

Norway, and especially in Luxembourg, the situation is reversed:as effective individual consumption. In addition, the aggregate

the individual consumption index is lower than the index ofstatistics do not allow for detailed analysis of differences in The variations in the job growth rate (see graph 1.2.1) reflect an increase in the job growth rate in 2004 which means GDP per capita. In the case of the Netherlands, this differentialstandard of living within countries and refined comparisons the economic trend, namely: an acceleration in economic growth; the GDP growth rate is explained by the significant volume of collective consump-between countries. Comparative results of harmonized surveys rose from 1.5% in 2001 to 2.5% in 2002, 2.9% in 2003

tion by the public administration. In the case of Ireland, andof income in European countries provide a wealth of supple- the crisis of 1975-1985, during which period the and to above 4% in 2004.

even more so for Luxembourg, a very important factor is thementary information, especially regarding inequality and po- Luxembourg economy even lost jobs at certain times;

volume of investment per capita, which is 39% above the OECDverty (see chapters 1.12 and 3.12). this loss of jobs primarily hit the manufacture of basic

average in Ireland and 103% above that average for Luxem- iron and steel in the midst of a crisis during this period;

bourg (as regards the development of gross fixed capital for-

mation in Luxembourg, see also p. 118). a continuous increase in the job growth rate beginning

in 1984, peaking - with a rate of 4.1% - in 1990 and 1991,

before returning to less spectacular increases during the

economic slowdown of 1992-1996; this downturn was

accompanied by a major increase in the unemployment rate

Graph 1.2.1

Rate of change in employment and unemployment rate, 1971-2004

in %

6.0 5.0 4.0 3.0 2.0 1.0

0.0 -1.0

Employment growth rate Source: STATEC, Employment Administration Average unemployment rate N.B.: unemployment rate = official ADEM rate

Graph 1.2.2 Graph 1.2.3

Economic and employment growth, 1985-2004 Productivity and employment growth, 1985-2004

Average annual GDP growth (in %) Average annual productivity growth in % (GDP/employment)

0. 3.5

IE IE

5.5 LU  

5.0 3.0 4.5

4.0 2.5 FI

PT

3.5

JP

3.0 2.0 SE

AT LU

UK

2.5 FR

2.0 1.5 IT EL

DK EU US

1.5 DE BE ES

NL

1.0 1.0

It should also be noted that the Luxembourg economy conti-The comparison of Luxembourg with the other OECD countries Average annual employment growth (in %)

Average annual employment growth (in %)

nued to create many jobs in 2002 and 2003, in contrast with(see graph 1.2.4) shows the key contribution of services -  Source: European Commission (AMECO) most other European countries experiencing economic diffi-principally traded services - to the significant job growth in

culties.This development, favourable all in all, did not howeverLuxembourg.This only confirms a fact that was already well

offset the increase in the unemployment rate, which rose fromknown. However, the comparative graph 1.2.4 indicating the Graph 1.2.4

3.0% in January 2002 to 3.7% in January 2003 and to 4.4% contribution of various economic sectors to job growth shows  Contribution by economic sector to job growth, 1990-2002

in January 2004. In spite of an acceleration in the rate ofa frequently neglected aspect: Luxembourg industry lost fewer (annual average contribution in %)

increase in jobs, the unemployment rate continued to increasejobs from 1990 to 2002 than other countries such as Belgium, 4 in % 3.7 3.5

in 2004 – admittedly at a slower pace than during the two Germany,Austria, Japan, Sweden and the United Kingdom. 3

previous years – reaching 4.7% in January 2005. Generally,In some countries, job growth in services was also offset by  2 Total employment growth 2.0 1.9 1.5

the upwards structural trend in unemployment should bea reduction in manufacturing employment, and as a result the 1 0.1 0.3 0.6 0.1 0.2 0.4 1.0 0.5 0.4 0.7 0.6 0.1 1.4 1.4 1.4 1.5 1.3

noted. Even in a period of exceptional economic growth, likeoverall job growth was very weak (e.g. in Germany, the United 0

that between 1997 and 2000, unemployment did not fall to Kingdom and Japan) or even negative (e.g. in Finland and -1

the low level of the beginning of the 1990s (as regards theSweden). -2 -0.5 -0.4 -0.4 question of unemployment, see also pp. 123 and 127).

Traded services Other services Manufacturing Other industries Source: OECD STAN Database, 2004

3. Graph 1.3.1

GDP growth rate in terms of volume, 1985-2005

From 2001 to 2004: 1120..00 in % LEuurxoezmonb eourg reversal in the economic  6420....0000 85-91: 7.4% 92-96: 2.9% 97-00: 8.0% 01-05: 3.0%

8.0

cycle and recovery T-2r.e0 nd of gross value added in terms of volume, 1995-2003 S(eosutirmcea:t eS TfAo Tr E2 C0 ,0 E4 U,  Rfo Or Se Tc Aa Tst  for 2005)

Graph 1.3.2

in %

16.0 Manufacturing, including energy

14.0 Financial services

12.0 Other services, excluding financial services,

10.0 public administration and education

8.0 All sectors 6.0

4.0

2.0

0.0

-2.0

-4.0 Source: STATEC (National Accounts)

The majority of the world's economies saw exceptionally highThe peaks in total value added growth of the Luxembourg Nevertheless, the overall decrease was spectacular. The comparison of GDP growth rates in Luxembourg and in the economic expansion during the second half of the 1990s.economy correspond to the combination of exceptional value In 2002 and 2003, the value added growth of financial  eurozone between 1985 and 2005 calls for a few additional Economic activity grew by an annual average of 4% in theadded growth rates in financial services, as well as in other services became positive again, although the growth rate remarks:

United States from 1996 to 2000; the figure for Europe duringservices (trade, transport and telecommunications, business is not comparable to the veritable explosion in value added

that same period was 2.7%.This growth reduced unemploy-services, etc.) and in manufacturing. It should also be noted in 1997 and 1999-2000. In addition, the pace of growth Luxembourg economic growth is extremely volatile, that is, ment, which fell from 10.5% in the EU-15 in 1994 to 7.4% inthat in 1998, the significant decrease in the value added in manufacturing and non-financial services remained veryit is subject to major fluctuations; this is characteristic

2001. In the United States, unemployment began to decline ingrowth of financial services (banks) and manufacturing was solid in 2002 and 2003, but did not reach the peaks of 1997of an economy that is largely exposed to external shocks; 1992 (from 7.5%), falling to 4% in 2000.This growth generallyalmost entirely offset by the very solid growth rates of non- and 2000 (for manufacturing) and 1998 and 2000 (for non-

occurred in a non-inflationary environment: apart from 1995,financial services. financial services).The result was rather sluggish growth in the past 20 years, during periods of strong economic

the rate of increase in consumer prices never exceeded 2.5% (at least in comparison with 1997-2000), with no economic expansion (1985-1991 and 1997-2000), growth in Luxem- during the second half of the 1990s, neither in the US nor inLuxembourg also slowed when the economies on which it sectors providing the impulse necessary to drive growth  bourg was around 2.5 times higher than average growth

the EU-15 taken as a whole. depends began to slow in 2001.This trend was emphasized  rates up sharply. in the eurozone: between 1985 and 1991, growth averaged

by the poor state of the financial markets.A decline in the 3.1% in the eurozone versus 7.4% in Luxembourg; between Luxembourg largely followed this growth pattern. Its GDPvalue added of financial services (negative growth rate), which 1997 and 2000, 8% in Luxembourg versus 2.9% in the growth rate averaged 8% annually from 1997 to 2000.Thisis the driving sector of the Luxembourg economy followed in eurozone;

growth surge was largely, but not exclusively, driven by the2001. Banks and other financial organizations respond to lower