WAGE MODERATION POLICY IN GERMANY Ray Barrell, Bettina Becker and Sylvia Gottschalk NIESR. 17 th November 2003

Transcription

1 WAGE MODERATION POLICY IN GERMANY Ray Barrell, Bettina Becker and Sylvia Gottschalk NIESR 7 th November 2003 Abstract: In this paper we briefly discuss the current condition of the German economy and the proposals from the government in March 2003 to reform labour markets. These reforms involve various measures to reduce the generosity of benefits and to increase the incentives to work. In order to analyse these reforms we simulate a temporary reduction in German real wage growth and discuss how German output, inflation, and interest rate are affected, under two distinct ECB monetary policy rules. We discuss the possible forward looking reactions of the ECB to these reforms and argue that it would be wise to respond in a fully accommodating way. The sensitivity of the ECB to changes in European key macroeconomic indicators is tested by analysing a scenario involving a reduction in French, Italian and German wages. It appears that single country wage moderation policies increase the country's competitiveness relative to other European countries, and hence helps raise the output effects in that country. We also consider that interest rates are set according to pre- 999 rules, with monetary policy more responsive to German conditions. German output would respond more quickly to the labour market reforms, but the long run consequences are the same, so there are no sustained benefits from an independent monetary policy.

2 I. Introduction Countries in monetary unions no longer have access to interest rate policies to stimulate demand, and members of EMU have also signed a fiscal pact that should restrict their freedom to use this instrument. Allsopp and Vines (998) amongst others have argued that members of EMU really only have recourse to labour market policies to expand output in a sustained way. Labour market reforms that cut real wages and increase participation increase output and also reduce unit costs, increasing the competitiveness of the economy. Outside a monetary union a central bank can give wage bargainers a clear signal that it will use an accommodating monetary policy if there is labour market reform, and this will speed up the process of adjustment and hence increase the net present value of the benefits to be reaped from reform. A country acting on its own with its own monetary authority, as in Germany before 999, might have the right incentive structure to engender labour market reform. A country within monetary union has a less well designed incentive structure to induce labour market reform, and hence it may be more difficult to achieve. We discuss these issues in relation to recent German policy initiatives. A country that does undertake labour market reform whilst not having its own monetary policy can still reap benefits, as the example of the Netherlands in the 980s and 990s shows. Barrell and Genre (999) discuss the Dutch competitive devaluation by wage moderation, and suggest that many of the gains from this reform came through increased competitiveness. These were especially important in relation to major trading partners such as Germany, against whom the exchange rate was pegged. However, it is clear that wage moderation is not a zero sum gain, and indeed, coordinated wage moderation can bring gains to all. The gains come from increased competitive vis-à-vis non-emu members, from a substitution of labour for capital because wages have fallen and from an expansion in overall aggregate demand. In the short run, of course, it may be the case that coordinated wage moderation leads to smaller gains for participants. It is easier for the central bank to respond to coordinate policies, and this may in turn make them more worthwhile. We begin by discussing recent developments in the German labour market and in policy initiatives toward that market. There have been a number of proposals that are designed to reduce the sustainable level of unemployment using policies toward nonwage labour costs and wage moderation and we detail them. We conclude our introduction by discussing how we might implement these policies on our model, NiGEM. In the second section we discuss production and price setting in our model, and also look at the determination of sustainable unemployment within the framework of our wage price system. Any labour market policy might impact on prices and inflation, and hence its effects are mediated by monetary policy. We discuss the 2

3 determination of interest rates in the last part of the second section of the paper. We then look at the impacts of wage moderation policies in Germany and in the Euro Area under current monetary arrangements, both with an active response to lower inflation and with a more accommodating policy that explicitly expands demand in response to the labour market developments. We also compare these effects to the impacts of wage moderation in Germany when monetary policy was operated by the Bundesbank. Our conclusions are that a pro-active response by the ECB might encourage wage moderation policies, and that there have been some costs to the efficiency of implementing policies in Germany as a result of moving to European Monetary Union.. The German Labour Market In light of the structurally weak condition of the economy and the labour market in particular, the German government has increasingly acknowledged the need for reform. Over the last few years, there have been various attempts to find strategies to foster employment and the competitiveness of the economy. However, clear signs of improvement have been slow to emerge. The standardised unemployment rate has been rising on a quarterly basis since 200, and the number of unemployed rose to more than 4.7 million in spring 2003, the highest in five years. The economy slid into recession in the third quarter of 200, with the downturn largely driven by a collapse in investment and the rapid depletion of inventories. The modest recovery that followed was led by exports, but only marginal economic growth was recorded in 2002, and the pace of growth was the slowest since 993. Investment has been further depressed by the collapse in stock markets last year, and the restructuring of the banking system, which may have contributed to slower credit growth, represents a downside risk to the prospects for investment and growth in Germany. The significant deterioration in public finances led to the activation of an excessive deficit procedure by the European Commission against Germany in November 2002, as the budget deficit widened to well above the ceiling of 3 per cent of GDP. Since the first quarter of 2003, the German economy has been in recession for the second time within less than ½ years, and the government s most recent estimates foresee a budget deficit of more than 3 per cent of GDP for the third consecutive year in 2004 (Bundesregierung, 2003). Standardised unemployment stood at 9.4 per cent for the fifth consecutive month in September 2003, up from an average of 8.2 and 7.8 per cent per annum in 2002 and 200, respectively. Figures for the West and the East continue to differ significantly. Unadjusted unemployment in East Germany exceeded that of the West by over 0 percentage points in the first seven months of this year, for instance. Between the second quarter of 999 and the last quarter of 2000, employment creation (persons employed) increased by 2.5 per cent. Measured in terms of hours worked the 3

4 employment gains were less than 0.5 per cent (OECD, 200). Whilst the first half of 200 saw employment peak, employment declined to around per cent below the peak in the second half of 2002 as economic activity deteriorated. However, the number of small hours jobs has increased in almost all sectors of the economy and subsidised short-time work nearly doubled in the first half of 2002 (OECD, 2003). With average working time decreasing across all sectors, the volume of total hours worked has fallen at a pace twice as fast as per-person employment since the first half of 200 (OECD, 2003). Compared to 99, average annual employment in 2002 increased marginally by 0. per cent, while total hours worked and hours per person in employment declined by 0.5 and 0.6 per cent, respectively (Federal Statistical Office, 2003). Most recently, employment in the second quarter of 2003 declined to a level about 2.2 per cent below the peak in 200, the lowest since the third quarter in 999, and the outlook for the labour market remains bleak. In 998 the newly elected government instituted round table talks between the social partners and the government, the Alliance for Jobs. Its main purpose was to develop a joint strategy for employment and competitiveness. Some informal guidelines were agreed upon, e.g., negotiated wage increases would not exhaust productivity gains and multiyear wage contracts would be preferred over one-year contracts. However, performance was disappointing, and growth slowed. Wage agreements rose significantly in the same year, after several years of wage moderation, and unit labour cost had begun drifting upwards even before the new wage agreements. March 2003 marked the collapse of the Alliance for Jobs. The German labour market has increasingly suffered from skill mismatches and shortages, resulting in companies in a diverse group of sectors not being able to fill vacancies. Between 99 and their peak since reunification in 2000, the average number of vacancies rose by 42.3 per cent per annum and has since fallen steadily to a level 6.4 per cent lower in the third quarter of 2003 compared to the average in 99. Legislation enabling immigration of qualified foreigners has not had the intended impact on aggregate unemployment figures. Moreover, since the implementation of the Job-AKTIV-Law in 2002, whereby the number of unemployed includes only those who are actively seeking work, unemployment figures have underestimated the rise in the non-working population as a number of people have withdrawn from the labour force. This contributed to the contraction of the labour force in 2002 (OECD, 2003). After the 2002 general elections, the new government merged the labour and economics ministries into one super-ministry in a move to address more effectively the reform processes needed to revive growth and significantly reduce unemployment. One of the new ministry s first tasks was to implement the labour market reforms 4

5 proposed by the Hartz commission. Key proposals were to reduce current unemployment benefit payments, speed up unemployment registration, increase temporary work opportunities through regional employment agencies, encourage under-the-table work to be brought into the tax net through tax incentives, and encourage greater labour mobility. The reforms Hartz I and Hartz II were written into law at the beginning of These include setting up Personal Service Agencies (PSA) at all employment agencies, increased support for self-employment ( Ich-AGs ), expansion of the low wage sector ( mini jobs ), and the job floater programme Kapital für Arbeit (capital for work). The objective of PSA is to make the process of entering the job market easier for the jobless and reduce employers reluctance to recruit. However, expectations have not been met so far. While original Hartz proposals suggested 500,000 jobless be transferred to PSA, 2,000 unemployed people were transferred by September 2003, and the government now expects only a total of 50,000 places both this year and next (Gemeinschaftsgutachten, 2003). In contrast, the financial support offered to jobless who enter self-employment has been sought after more frequently than expected, with just under 62,000 Ich-AG s established in the first nine months since the implementation of the measure in January 2003 (Gemeinschaftsgutachten, 2003). However, since eligibility for the programme does not require a promising business concept, it remains to be seen how many Ich-AGs will be able to survive over the medium term. The only demand side programme, Kapital für Arbeit, whereby companies receive the option to take out a loan over up to 00,000 for each permanently employed jobless person, has so far been little successful. Only about 8,000 jobs were created between November 2002 and September Furthermore, whilst 930,000 new mini jobs were counted between April and June 2003 as a result of the expansion of the low wage sector, this figure is likely to overestimate the programme s success. As the number concerns jobs rather than persons, it may include many double counts (Gemeinschaftsgutachten, 2003). The government has frequently been criticised for implementing the reforms in a much weaker version than originally suggested by the Hartz commission, and overall success of the reforms has been moderate so far. It is planned that two more proposals be written into law in January and July 2004, Hartz III and Hartz IV, respectively. Whilst these two proposals, which involve the reform of the Federal Employment Agency and the merger of unemployment and social security benefits, were approved by the Bundestag on 7 October, implementation of Hartz IV now depends on approval by the opposition-dominated Bundesrat, and in case of rejection a compromise to be suggested by the mediation committee may be agreed upon in December. 5

6 The Hartz measures are part of the Agenda 200, a paper demanding urgent reform and intended to provide new incentives for labour, consumption and investment, which was released by Chancellor Schröder in March With respect to the labour market, a large part of the agenda consists of measures to lower non-wage labour costs by substantially reducing entitlements. Contributions for unemployment insurance, pensions, healthcare, and nursing cover amount to nearly 42 per cent of gross wages in Germany and are widely seen as the major competitive burden on the economy. Hartz IV has been designed as one means of addressing this problem. In addition, average contribution rates for statutory healthcare are to fall from 4.3 per cent to 3.6 per cent from 2004 and below 3 per cent from Payment of unemployment benefits for people aged up to (more than) 55 years will be restricted to twelve (8) months. Thereafter, only welfare benefits are to be paid rather than long-term unemployment benefits, which are usually significantly higher. It is planned that these changes will take effect from January They return to one of the main proposals in the preliminary report of the Hartz commission, which had later been dropped due to union pressure. In order to plug a shortfall of an estimated 8 billion in the state pension scheme without increasing pension contributions, latest government plans foresee a freeze of pensions next year for the first time in postwar German history. From April 2004, pensioners will be required to pay their contributions to the statutory old-age care scheme in full, part of which is currently covered by the pension funds. Without corrective measures, contribution rates would have had to rise from currently 9½ per cent of gross wages to 20.3 per cent. The agenda also suggests first measures to loosening Germany s job protection. In light of the excessively high unemployment figures, Germany s biggest trade unions have indicated limited willingness to compromise on small changes. It is now planned that from the beginning of next year, small companies with currently five employees will be allowed to employ up to five full-time equivalent employees on fixed-term contracts without the consequence of the other employees becoming eligible for full job protection. There are also signs that the rigid wage bargaining system may be modestly modified so as to allow for some greater local flexibility. Successful implementation of the reform measures could enhance the efficiency of labour market policy and slow the pace of the increase in the unemployment rate. Collectively-bargained hourly wages increased by some 3 per cent in 999, but compensation per employee increased just below 2 per cent, and compensation per hour worked increased above.5 per cent. This was due to reductions in non-wage labour costs and faster growth in part time employment led to lower increases in per capita real wage. In 2000 growth in collectively bargained wages came down to 2 per To persons aged above 55 the new law would apply only from

7 cent, and increases in real wage per employee were held at.25 per cent. In addition, wage negotiations led to mostly 2-year wage contracts with average collectively bargained wage increases of around 2 per cent for 200 (OECD, 200 and 2003). The negative wage drift was small, mainly due to the stagnation in the growth of small hours jobs. Average wage growth remained about 0.5 per cent below consumer price inflation. The wage round concluded in spring 2002 implied wage increases of 3-3½ per cent in the first year of the contract period. Increases in the second year were somewhat lower. Accordingly, wages rose in the latter half of 2002 (OECD, 2003). Moderate increases were implied by this year s wage round, amounting to significantly less than 2 per cent in the wholesale and retail trade sectors for the whole of the 2-year contract period, for instance (Benner et al., 2003). Most recently, hourly wages in the second quarter of 2003 as well as collectively-bargained monthly wages in July 2003 were 2.6 per cent above their previous-year level (Federal Statistical Office, 2003)..2 Wage Moderation and NiGEM Our objective is to evaluate in outline the potential impacts of the policies discussed above, and we analyse how potential output and its rate of growth are affected by a policy of wage moderation over an extended period of time. An endogenous shock is applied to German wage growth and two simulations are then carried out. The wage moderation policy is incorporated in the ECB s inflation forecast and the short-run interest rate is adjusted accordingly. This forward-looking ECB monetary policy, whereby the monetary authorities use their forecast, reduces short-run interest rates at the date of the shock. However, it is possible for the ECB to be more accommodating whilst maintaining price stability, and we analyse this by assuming that the ECB shifts its nominal target by increase in the level of European real GDP. Such a policy maintains price stability in the medium term which is the ECB s overriding objective, whilst ensuring that the benefits of labour market reform feed through more quickly. Such a response might encourage social partners to cooperate in much needed reforms. A similar shock in wage growth is applied to three major European economies, France, Germany and Italy. We compare the more accommodating ECB policy with the respective case where wage moderation is pursued only in Germany. We then ask what would have happened if European interest rates were still set under the pre-999 arrangements. We found that in the short-run German output and employment growth are much higher when German monetary policy is independent. Inflation and interest rates are also lower. However, the longer term benefits are essentially the same whether or not Germany has an independent monetary policy, and if it still had one it 7

8 would not be able to reap the significant gains that should be available from a single currency rather than just a set of fixed exchange rates. II. Modelling the Economy The analysis of a German wage moderation policy was carried out in a large scale, rational expectations, New Keynesian macro-econometric model: the National Institute global model, NiGEM. It is an estimated model of the world economy firmly grounded in economic theory. The underlying structure determines the level of output and the trajectory for prices, with equilibrium correction mechanisms pulling the economy toward its long run, albeit slowly. 2.. Production and price setting For each country we have an underlying CES production function which constitute the theoretical background for the specification of the factor demand equations and provide a measure of capacity utilisation which then feeds into the price system. 2. The CES production function, with constant returns to scale, can be written as follows () Y = γ [ δ ( K) ρ + ( δ )( Le tech ) ] ρ / ρ This is associated with a unit cost function (2), (2) σ UTC = (/ γ ) * (( δ ) * user σ + (( δ ) σ ) * wage σ e ( σ )* tech ) (/( σ )) The associated steady state price equation involves a mark up over unit costs, where the dynamics of the mark up will depend upon capacity utilisation (CU), for instance. If we can recover the parameters of the production function we can write out our unit cost function and our price equation (3). (3) P = θ * UTC * e π * CU The elasticity of substitution, = (/(+ )) can be retrieved from a labour demand curve, as can the coefficient on labour augmenting technical progress. Once we have these parameters the distribution parameter and the scaling parameter can be calibrated. The steady state version of the price equation we use in this analysis puts the user cost of capital, capacity utilisation and the mark up at trend values. We derive labour demand as a function of real wages, technical progress and output (Barrell and Pain, 997). The mark-up adjusted real wage (m*(w/p)) can be set equal to the marginal product of labour derived from the production function, and it can then be rearranged to give a labour demand curve conditioned on output, the real wage and technical progress, and we use this and the associated capital demand equation in the model. 2 Cobb-Douglas imposes a unit elasticity of substitution and embodies neutral technological progress. This would be too restrictive for our purpose. 8

9 2.2. Labour market We assume that labour markets embody rational expectations, at least where we have evidence that bargainers use forward expectations of future inflation (Barrell and Dury (2003) and Anderton and Barrell, 995 discuss our work). We estimate the natural rate of unemployment within a stylised version of the bargaining framework of Layard et al. (99, Chap. 2). Firms are assumed to have the right to manage, determining employment according to their labour demand curves. Therefore bargaining only takes place over wages, the outcome reflecting the relative strength of unions versus employers. The NAIRU is an equilibrium concept corresponding to the rate of unemployment that would prevail were the endogenous wage and price variables at their equilibrium levels (Barrell et al., 993). The wage equation models in a stylised way the determinants of the bargaining outcome by making wages dependent on the prevailing unemployment rate (U), as well as on average labour productivity. Clearly, many other factors such as replacement ratios, the degree of unionisation or the existence and level of minimum wages influence the bargain. In this analysis they are encapsulated in the intercept, α, and we can analyse them by shifting the intercept. Factors that are subject to likely reform along the lines recently suggested by Mr Schröder, such as the generosity of benefits, the structure of bargaining and job protection are also incorporated in the intercept term. We may write the wage equation as: (4) ln( Wt / Pt ) = α + ln( Yt / Lt ) βu t An estimate of the NAIRU can now be obtained by solving the price equation for the (log) real wage, inserting the result into the equation for the real wage and solving the latter for the unemployment rate, assuming that capacity utilisation is at trend: NAIRU (5) t Y = { ln β L + /( σ ) log{( t t λ t γ / θ ) + a ( σ /( σ )) * log( /( δ ) δ * ( user / P ) ( σ ) σ ( σ ) }]} In a CES world the sustainable rate of unemployment depends on the user cost of capital, the mark up of prices over costs as well as on the parameters of the production function and the intercept in the wage bargaining equation. If we had a Cobb Douglas production function we would have a differently specified cost function, and hence we cannot take this as a special case of CES. If we wish to shift the Nairu in the model we have a number of levers to do so, including the intercept in the price equation, which is the mark up, and the real rate of interest. However, the former is influenced by competition policy whilst the latter, to the extent it can be changed, is influenced by the longer term stance of fiscal policy. As we are interested in active 9

10 labour market policies as described in the previous section we choose to shift the level of the wage equation, reflecting policies that reduce sustainable level of unemployment through greater labour market participation Monetary policy rules We assume that the monetary authorities use the interest rate to target the price level or the inflation rate in the long term. The speed of response of the authorities affects the properties of the model. In our forward-looking world the expectation that interest rates would be lower would mean that the exchange rate would decline now. This would improve competitiveness in the short run and would raise demand. This would eventually increase prices as compared to where they would have been. A typical policy for a central bank may be to target some nominal aggregate such as nominal GDP or the money stock, which may rise in line with nominal GDP in the long run. A standard monetary policy rule would be to change the interest rate according to some proportion of the deviation of the targeted variable from its desired path. The ECB has adopted a combined rule with 2 pillars. A combined policy of nominal aggregate targeting and inflation rate targeting would, with P indicating a price level and Y indicating output, then give (6) rt = γ (log( Pt Yt ) log( P * t Yt *)) + γ 2 ( log Pt log Pt *) The policy rules on the model centre around this type of rule where the inflation rate being targeted is the Consumer Price Index (CPI). An asterisk denotes a target and all right hand side variables are in logs. We choose the combined rule as our default monetary policy rule, in part, because it describes what we think central bank behaviour should be, but also because it represents the mixed framework that is used in Europe by the European Central Bank (ECB). The addition of the inflation component to the rule is particularly important when analysing shocks, as we illustrate below. III. Wage moderation policy in Germany We can ask whether and to what extent co-ordinated labour market and monetary policies can raise potential output in Germany and the Euro area. We are interested in investigating how potential output is affected by a policy of wage moderation over an extended period of time. We are also concerned with the effects of this change on employment, the price level and the interest rate. Wage moderation is implemented in NIGEM by changing the intercept (or residual) on the real wage per person hour equation (4 above) by one percent whilst leaving the equation endogenous, and we do this first for Germany under two different policy reactions. If labour market reforms are made in isolation then the effects may come through more slowly and hence the 0

11 incentive to undertake them is reduced as the net present value of the benefits is lower than if policy accommodates the labour market initiatives. Figure : Germany Real wages Active and accommodating monetary policies % difference from baseline Quarters Active monetary policy Accommodating monetary policy Figure 2: Germany Total employment Active and accommodating monetary policies % difference from baseline Quarters Active monetary policy Accommodating monetary policy After the reduction in the wage equation intercept of per cent the real wage 3 subsequently declines to just over 2.5 per cent difference from its baseline, before beginning to recover about two and a half years after the shock due to increased 3 Real wage per person per hour. RWAGE=(COMP/(EE*HOURS))/CED, where COMP is total compensation, EE total employment, HOURS, the number of hours worked per person and CED, the Consumer Expenditure Deflator.

12 demand for labour, as can be seen from figure. Wage moderation induces an increase in the number of persons employed, as can be seen from figure 2. In the long run, total employment is more than.5 per cent higher and real wages are almost.5 per cent lower following the wage moderation policy. The increase in capacity output puts downward pressure on prices, and the ECB has to decide how to respond. When we analyse active policy responses we assume that the central bank is forwardlooking when it makes its forecast of inflation, but does not alter its other pillar, the stock of money. Hence wage moderation is incorporated into the ECB s inflation forecast and the short-run interest rate is adjusted accordingly. This reaction will help stabilise inflation, but the labour market reform will mean that it will be below target, and prices will be increasing less rapidly than they otherwise would have done. The ECB has the objective of maintaining price stability in the medium term and it is failing to do so in our active response experiment. In order to do this it would have to shift the nominal target in the two-pillar strategy and we describe this as an accommodating policy. 4 The shift required in the nominal target is determined by the long-run increase in European real GDP, which we can see from figure 3 is 0.5%. Hence we can repeat our experiment with an endogenous shock of -% applied to the German real wage, along with an exogenous shock of 0.5% to the European nominal target. Figures 3 and 4 compare the active and accommodating policies. Figure 3: Output in Germany and the Euro Area Active and accommodating monetary policies % difference form baseline Quarters Euro Area - active policy Germany - active policy Euro Area - accommodating policy German - accommodating policy A shock in the ECB s nominal target has positive impacts on both the German and European real output in the short to medium term, but no long run impact. Comparing 4 Henceforth we will refer to the forward-looking policy without the shift in the nominal target as the ECB s active policy, versus the accommodating policy of shifting the target. 2

13 output growth in Germany and the EU under active and accommodating monetary policies shows a higher rate of growth under the latter in the first few years. As a consequence of the more positive policy response European output reaches its long run equilibrium more quickly and then stabilises at that level. German output also rises more rapidly and reaches its long run equilibrium level in the medium term. The consumer price index under active and accommodating monetary policies follow analogous paths as can be seen from figure 4. With accommodating policy the Euro Area price level returns to target in the long run while remaining about 0.7 per cent below base without the shock to the nominal target. European interest rates are slightly lower in the short run when the ECB pursues the objective of price stability by shifting the nominal target accordingly, while they return to base in the medium term under both scenarios. Clearly the net present value of a labour market reform is higher if the central bank responds more positively as the effects come through more quickly. Figure 4: The Price Level in Germany and the Euro Area Active and accommodating monetary policies % difference from baseline Quarters Europe - active policy Germany - active policy Europe - accommodating policy Germany - accommodating policy IV. Wage moderation in the Euro area. One potential cost of moving to EMU is that the central bank will respond rather less to labour market reform that is undertaken in one country at a time than would a central bank oriented only to developments in that country. Hence with either an accommodating or an active policy response there is a case for acting in concert rather than in isolation. We may analyse this by applying similar shocks to the major economies of Europe and evaluating the responses. Figures 5 and 6 show how a policy of wage moderation in three major European economies in the Euro area affects output, inflation and interest rate in Germany and in Europe. We consider the monetary policy where we have full accommodation which involves shifting the 3

14 European nominal aggregate to ensure price level stability. The main difference between the results of this section and the last is that we see larger variations in output and inflation. Figure 5 shows that European output growth is much stronger than under the respective policy scenario in the previous section, where only German wages were being reduced. Output growth in Germany is higher than European output growth in the short term. Figure 5: Output in Germany and Europe accommodating monetary policy German and European output % difference from base quarters German Output Euro Area Output Figure 6: Inflation in Germany and the EU accommodating monetary policy percentage points difference from baseline quarters German Inflation Euro Area Inflation It is also possible for us to compare the gains that might be made from German labour market reform undertaken in isolation to those that might come from coordinated 4

15 wage moderation in the Euro Area. Some of the gains to wage moderation come from increased competitiveness with respect to trading partners within the Area and outside it, and if wage moderation is coordinated then some of these competitive devaluations by wage moderation gains might be lost. However, the expansion in activity in the Euro Area itself might be beneficial, as the fall in the real wage in a number of countries would stimulate the level of activity and employment in the medium term. It is clear from figure 7 that the initial gains from labour market reform when it is coordinated with other countries are slightly smaller than when it is undertaken in isolation. However, after a few years the gains to the Area as a whole are sufficiently large that they overcome this small competitiveness based losses, and the longer term impact of coordinated policies are 50 per cent higher than policies undertaken in isolation. Figure 7: The Impact of Coordinated and German wage policies on German output (ratio of coordinated to German policy) Ratio Quarters German output : a coordinated change in European labour markets compared to German labour markets only change V. German targeting We finally carry out a simulation where we assume that European interest rates are set under the pre-999 arrangement, where Germany had an independent monetary policy. The same structure of labour market reforms is assumed to be introduced, with the intercept on the equation for the German real wage per person hour being reduced by the same amount as in our first EMU based experiment. The German authorities use the same feedback parameters but target German aggregates. Hence a shock that 5

16 reduces inflation in Germany rather more than it does in the Euro Area as a whole will engender a different policy response under German targeting. Figure 8 Differences in Output and Inflation in response to a German labour market shock under German versus European targeting Percentage points difference between German and European targeting quarters Output: difference between German and European targeting Inflation: difference between German and European targeting Inflation under German targeting: difference from base Figure 9 Interest rates in response to a German labour market shock under German and Euro Area targeting percentage points quarters Difference in interest rates between German and Euro Area targeting 6

17 A comparison of key German indicators under the active ECB monetary policy described in section II and German targeting provides interesting insights. 5 In the short to medium term, German macroeconomic indicators of output and employment are better under German targeting and active monetary policy. German output in the short term is much higher when German monetary policy is independent, as can be seen from figure 8. However, the longer term benefits are the same whether or not Germany has an independent monetary policy, and if it still had one it would not be able to reap the significant gains that should be available from a single currency rather than just a set of fixed exchange rates. Inflation is also higher, as can be seen from figure 8, but in both cases German inflation is well below its baseline level in the early years of the experiment. Interest rates are much lower as a result of the decision to implement German targeting because German inflation falls rather more below its baseline than does European inflation, and the ECB is required to react to European aggregates. VI. Conclusion We have investigated the effects of a permanent reduction in real wages in Germany on output, employment, interest rates and inflation. In particular, we were interested in analysing how the ECB responds to this wage moderation policy in one of the largest European economies. If the monetary authorities anticipate that wage moderation will reduce inflation, increase employment and hence output, the shortrun interest rate declines at the time of the shock. The positive impacts of the wage moderation policy on both the German and European real output growth are enhanced when the ECB policy is fully accommodating, shifting its nominal target by the growth rate of the European GDP so as to maintain price stability. If the Bundesbank were still running monetary policy in Europe the German economy would respond more quickly to the policy change, but in the long run the impact on output would be the same. However, inflation and hence interest rates tend to be higher in the short to medium term if Germany were in control of monetary policy. 5 For the purposes of this paper, the discussion is kept brief here. Current research at NIESR looks at the costs and benefits of an independent German monetary policy in more detail. 7

18 References Allsopp, C. J., and Vines, D., (998) Macroeconomic Policy after EMU Oxford Review of Economic Policy vol 4 no. 3 pp -23 Anderton, R. and Barrell, R. (995), The ERM and Structural Change in European Labour Markets: A study of 0 countries, Weltwirtschaftliches Archiv, Band 3, Heft. Barrell R., and Dury, K. (2003) Asymmetric labour markets in a converging Europe: Do differences matter? National Institute Economic Review No. 83, January 2003 Barrell, R. and Genre, V. (999), Employment Strategies for Europe: Lessons from Denmark and the Netherlands, National Institute Economic Review, April. Barrell, R. and Pain, N. (997), Foreign direct investment, technological change, and economic growth within Europe, Economic Journal, 07, pp Barrell, R., Pain, N. and Young, G. (993), Structural Differences in European Labour Markets, NIESR Discussion Paper No. 46. Benner, J., Borbély, D., Boss, A., Kuhn, A., Meier, C.-P., Oskamp, F., Scheide, J. and Schmidt, R. (2003), Leichte Belebung der Konjunktur in Deutschland, Die Weltwirtschaft, Heft 3, forthcoming. Bundesregierung (2003), various news releases, Federal Statistical Office (2003), various data releases, Gemeinschaftsgutachten der Arbeitsgemeinschaft deutscher wirtschaftswissenschaftlicher Forschungsinstitute e.v. (2003), Die Lage der Weltwirtschaft und der deutschen Wirtschaft im Herbst 2003, Munich (cited as Gemeinschaftsgutachten ). Gern, K.-J., Meier, C.-P. and Scheide, J. (2003), Higher economic growth through macroeconomic policy coordination? The combination of wage policy and monetary policy, Kiel Institute for World Economics Discussion Paper No Layard, R., Nickell, S. and Jackman, R. (99), Unemployment: Macroeconomic Performance and the Labour Market, Oxford University Press. NIESR (2002), The World Model Manual, mimeo, April. OECD (200), OECD Economic Survey on Germany, May. OECD (2003), OECD Economic Survey on Germany, January. 8