52003DC0309

REPORT FROM THE COMMISSION TO THE COUNCIL AND THE EUROPEAN PARLIAMENT - Experimental application of a reduced rate of VAT to certain labour-intensive services [SEC(2003) 622]

CONTENTS

SUMMARY

1. Origin and objectives of the experiment

2. The underlying economic mechanism

2.1. Employment: increasing demand by reducing prices

2.2. The black economy

3. The economic environment in the trial period

3.1. Rates and scale of VAT

3.2. General economic conditions

3.3. Economic indicators for specific sectors

3.3.1. Renovation and repair of private dwellings

3.3.2. Other sectors

4. Evaluation of the experiment

4.1. Evaluation method

4.2. Lower rates of VAT not fully reflected in prices

4.3. Expansion promoting demand

4.4. No substantial impact on employment

4.5. A measure with purely relative impact

4.6. Black economy: evaluation remains difficult

5. General conclusion

SUMMARY

Origin of the experiment

In December 1998 the European Council concluded that promoting employment, economic growth and stability was one of the four subjects that most concerned Europeans, and called for rapid and effective action. The Presidency conclusions invited the Commission "to allow those Member States who so desire to experiment with reduced VAT rates on labour intensive services which are not exposed to cross-border competition."

Council Directive 1999/85/EC of 22 October 1999 amends Directive 77/388/EEC as regards the possibility of applying on an experimental basis a reduced VAT rate on labour-intensive services. The objectives of this experiment were to increase employment and reduce the black economy.

The experiment was strictly limited in time and concerned only the services described in the new Annex K added to Directive 77/388/EEC. Nine Member States chose to participate in the experiment.

Services and Member States concerned

1. Small repair services: //

- Bicycles // B, L, NL

- Shoes and leather goods // B, L, NL

- Clothing and household linen // B, EL, L, NL

2. Renovation and repair of private dwellings // B, E, F, I, NL, P, UK

3. Window cleaning // F, L

4. Domestic care services // EL, F, I, P

5. Hairdressing // E, L, NL

Because of the experimental nature of the measure, Directive 1999/85/EC stipulates that participating Member States must draw up a detailed assessment of its impact in terms of job creation and efficiency. The Commission is required to submit a global evaluation report to the Council and the European Parliament accompanied, if necessary, by a proposal for appropriate measures for a final decision on the VAT rate applicable to labour-intensive services.

Monitoring the experience; common frame of reference for evaluation

The Commission was in charge of implementing and monitoring evaluation of the experiment. In July 2002 a preparatory meeting with the participating Member States was held at which there was an exchange of views on the evaluation methods to be used and the difficulties encountered. In October 2002 another meeting attended by all the Member States provided an opportunity for the participating Member States to present and discuss the results of their assessment reports.

The Commission's evaluation incorporates the evaluation system discussed at the preparatory meeting in July 2002. Each stage of the economic mechanism is examined in detail. This common frame of reference is applied to each Member State's assessment report.

Aiming for job creation: a very specific economic mechanism

The link between reduced rates of VAT and job creation is not a direct one. It is based on the fact that if the final price of a product falls sufficiently, demand for that product will rise, all else being equal.

For the mechanism to work, a reduction in VAT must lead to a fall in the price charged to the consumer of the service concerned. The price must fall sufficiently to generate increased demand for the service. Increased demand should, in turn, lead to increased production. The increased production must then be covered by hiring new staff, and not by raising productivity (which would mean producing more with the same number of employees) or increasing the working time of the staff already employed.

If the fall in the VAT rate is used by the vendor to increase his profits, the mechanism can no longer work as intended. In this case the reduced rate of VAT acts as a subsidy for a particular sector by reducing a tax which does not represent a cost to the firm. To avoid this undesirable outcome, Directive 1999/85/EC stipulates that there must be a close link between the lower prices resulting from the rate reduction and the foreseeable increase in demand.

Aiming to reduce the black economy

This measure can only have an impact on firms which already have legal status but which operate partly in the black economy. The problem with evaluating that impact is that of measuring activity which, by definition, evades the obligation to submit returns.

For businesses whose very existence is not declared, reducing the rate of VAT does not offer an incentive to leave the black economy, since none of their production is declared and they pay no production-related taxes. VAT is not the main reason why they operate in the black economy, but rather the fact that this means paying no taxes or social security contributions at all.

Lower rates of VAT not fully reflected in prices

When they conducted price surveys, Member States found that reduced rates of VAT are never fully reflected in consumer prices. Part of the VAT reduction is used to increase the margins of service providers.

Renovation and repair of private dwellings seems to be the sector in which service providers most readily pass on the VAT reduction. This may be explained by the average scale of expenditure involved. Nevertheless, the reduction is often only temporary.

In hairdressing, where a part of the reduced rate of VAT was passed on immediately, consumer prices were subsequently increased by more than the rate of inflation. Thus the application of a reduced rate only temporarily curbed the customary price increases.

In certain cases, most notably that of bicycle repair, service providers refused to apply the reduced rate, arguing that the measure was too complex.

A context of economic expansion

The reduction of VAT rates was applied at a time when growth had been vigorous for some years and there was a downward trend in both unemployment and inflation. The overheating of the economy led to renewed inflation in 2000.

The increase in demand observed can be explained by the increase in households' disposable income. In addition other measures, particularly tax measures, were introduced and contributed to the growth of demand independently of the VAT rate reduction measure.

The Member States' reports do not allow the effects of growth, particularly vigorous at the time, to be distinguished from the possible effects of VAT rate cuts being partially passed on in prices.

No substantial impact on employment

Some Member States took a pro-active approach to employment and assistance, concentrating on certain sectors, particularly construction.

When the sectors tried to produce more to meet increased demand, they encountered difficulties with recruitment, or resorted to overtime.

Under these circumstances the Member States' reports do not identify solid evidence of reduced VAT rates having an impact on employment.

Only France and Italy attribute the creation of a relatively large number of jobs in the home renovation and repair sector to the reduction in VAT rates. However, their approach failed to take a number of relevant factors into account.

The figures presented in the French report indicate that the cost of such job creation is very high, at about EUR 89 000 per job per year.

Measure's effectiveness entirely relative

Empirical studies of the effects of reducing VAT rates show that such measures are never the most effective and that the cost to the budget is high in relation to any impact the measures may have on the economy.

Using its macro-economic QUEST model, the Commission conducted a simulation exercise especially for this evaluation report. Reducing VAT rates is more costly in budget terms than other measures that directly target labour costs. These results are entirely consistent with the conclusions of previous studies.

Calculated for the EU as a whole, reductions in labour charges create 52% more jobs than reductions in VAT rates for the same cost to the budget.

Black economy: evaluation remains difficult

Member States' reports did not succeed in overcoming the difficulty of measuring the black economy. It was not possible to demonstrate that the measure had contributed to reducing the black economy. Most of the results are no more than hypotheses unsupported by evidence of any causality between the measure and any decrease in the black economy.

The Commission will take account of the results of this experiment in the review of the scope of reduced rates provided for in Article 12(4) of the Sixth Directive. The review will be guided by the objectives of the new VAT strategy, namely to modernise the common system of VAT, simplify it, strengthen it and apply it in a more uniform fashion.

1. Origin and objectives of the experiment

In 1997 the Extraordinary European Council on employment of 20 and 21 November concluded that "the issue of employment is central to the concerns of Europe's citizens and every effort must be made to combat unemployment, the unacceptable level of which poses a threat to the cohesion of our societies." In its conclusions the Presidency invited every Member State to examine, without obligation, the advisability of reducing the rate of VAT on labour-intensive services not exposed to cross-border competition.

In the same year, the Commission Communication to the Council on "Job creation: possibility of a reduced VAT rate on labour-intensive services for an experimental period and on an optional basis" [1] stated that the best way to stimulate employment through tax measures was through a reduction of labour costs other than wages that was neutral in terms of public finances. It argued that a targeted reduction of social security contributions was probably a more efficient way of promoting job creation than a reduction in VAT rates, and pointed out that there were other tax instruments whose effectiveness in combating unemployment had already been demonstrated.

[1] SEC(1997) 2089 final.

The European Council of 11 and 12 December 1998 reaffirmed in the "Vienna strategy for Europe" that promoting employment, economic growth and stability was one of the four subjects that most concerned Europeans, and called for rapid and effective action. Point 35 of the Presidency conclusions invited the Commission "to allow those Member States who so desire to experiment with reduced VAT rates on labour intensive services which are not exposed to cross-border competition."

The proposal for a Council Directive of February 1999 [2] was a response to that invitation. Point 4 of the Explanatory Memorandum pointed out that such an initiative could, however, have a negative impact on tax neutrality and on the smooth functioning of the single market, and that the positive effects of VAT relief on job creation were not undisputed, particularly as regards the reduction in rates actually being passed on in consumer prices. It stated the Commission's view that the best way to promote employment was to maintain the policy of reducing all charges on the labour factor, particularly in the case of low-skill, low-pay labour.

[2] COM(1999) 62 final.

On 22 October 1999 the Council adopted Directive 1999/85/EC amending Directive 77/388/EEC as regards the possibility of applying on an experimental basis a reduced VAT rate on labour-intensive services.

Recitals (2) and (3) of the Directive state the objectives of this experiment, namely to increase employment and reduce the black economy.

The experiment was strictly limited in time and concerned only the services described in the new Annex K added to Directive 77/388/EEC.

The services concerned had to satisfy the following requirements:

a) they must be labour-intensive;

b) they must be, in the main, provided directly to final consumers;

c) they must be mainly local and not likely to create distortions of competition;

d) there must be a close link between the lower prices resulting from the rate reduction and the foreseeable increase in demand and employment.

Member States could apply a reduced rate of VAT to two, or in exceptional cases three, of the categories of services defined in Annex K. Member States wishing to apply the measure had to provide the Commission with all the information necessary to demonstrate that the conditions were met, as well as a precise description of the scope of the measure, the services concerned and the expected budgetary cost.

Because of the experimental nature of the measure, Directive 1999/85/EC stipulated that participating Member States must draw up a detailed assessment of its impact in terms of job creation and efficiency.

The Commission was required to submit a global evaluation report to the Council and the European Parliament accompanied, if necessary, by a proposal for appropriate measures for a final decision on the VAT rate applicable to labour-intensive services.

The subject of this report is the global evaluation of the experiment.

Nine Member States chose to participate in the experiment. The participating Member States and the services selected are listed in Council Decision 2000/185/EC of 28 February 2000 authorising Member States to apply a reduced rate of VAT to certain labour-intensive services in accordance with the procedure provided for in Article 28(6) of Directive 77/388/EEC. [3]

[3] 2000/185/CE.

Services and Member States concerned

1. Small repair services: //

- Bicycles // B, L, NL

- Shoes and leather goods // B, L, NL

- Clothing and household linen // B, EL, L, NL

2. Renovation and repair of private dwellings // B, E, F, I, NL, P, UK

3. Window cleaning // F, L

4. Domestic care services // EL, F, I, P

5. Hairdressing // E, L, NL

To monitor assessment of the experiment, the Commission sent a questionnaire [4] to the participating Member States setting out the main points necessary for a reliable evaluation of the measure. In July 2002 the Commission organised a preparatory meeting with the participating Member States, at which views were exchanged on the evaluation methods to apply, and the difficulties encountered by Member States in drawing up their assessment reports were discussed.

[4] See annex.

In October 2002 the Commission received all the reports from the Member States. During the same month a meeting organised by the Commission provided an opportunity for the participating Member States to present and discuss the results of their assessment reports with all the Member States and the Commission.

2. The underlying economic mechanism

The stated objectives of the experimental reduction of VAT on labour-intensive services were job creation and the reduction of the black economy.

2.1. Employment: increasing demand by reducing prices

The economic mechanism underlying the measure is based on the fact that if the final price of a product falls sufficiently, demand for that product will rise, all else being equal.

It is therefore essential for a reduction in VAT to lead to a fall in the consumer price of the service concerned. This price reduction should generate increased demand for the service. Increased demand should, in turn, lead to increased production. To increase production, jobs will be created, provided there are not other ways of increasing production.

Clearly, the link between reduced rates of VAT and job creation is not a direct one. The mechanism only works if certain conditions are fulfilled.

In particular, the reduction in VAT must be passed on in the consumer price by the vendor. Consumer prices must fall sufficiently to increase demand. The maximum possible price reduction is determined by the difference between the standard rate of VAT and the reduced rate applied. The increased production must be covered by hiring new staff, and not by raising productivity (which would mean producing more with the same number of employees) or increasing the working time of the staff already employed.

If the fall in the VAT rate is used by the vendor to increase his profits, the mechanism can no longer work as intended. In this case the reduced rate of VAT acts as a subsidy for a particular sector by reducing a tax which does not represent a cost to the firm. In practice the VAT is ultimately borne entirely by the consumer, who then, if prices do not reflect the reduction in VAT, bears the full cost of this subsidy to the sector. To avoid this undesirable outcome, Directive 1999/85/EC requires there to be a close link between the lower prices resulting from the rate reduction and the foreseeable increase in demand.

If the profit is used to hire extra labour, this indicates that there is demand which is not being met. This unsatisfied demand justifies margins of price increase.

In other words, the purpose of a tax on consumption is not to influence production costs.

In conclusion, the link between reduced rates of VAT and job creation is not a direct one. For the economic mechanism linking cut rates of VAT to job creation to work, a whole series of effects must occur.

2.2. The black economy

Businesses operating in the black economy can be divided into two categories:

i) those which operate in both the formal economy and the black economy and

ii) those businesses or individuals that operate entirely outside the formal economy.

The first category is thus made up of properly registered businesses with a VAT number, which pay their social security contributions and corporation tax and fulfil other tax obligations. A part of their production is, however, not declared. In the second category all production is carried out by undeclared producers, who do not, therefore, pay VAT, corporation tax or any other sort of taxes or social contributions.

Clearly, for the second category of producers operating entirely in the black economy, reduced rates of VAT are not a sufficient incentive to abandon their undeclared status. All their activity is undeclared and they pay no taxes in connection with it. The main reason for their operating in the black economy is not VAT but the total absence of any taxation or social security charges.

The measure can, therefore, only influence those businesses which have legal status but which operate in part in the black economy.

They may be encouraged to reduce their underground activity by a possible increase in consumer insistence on an invoice. For instance, in the case of renovation and repair of dwellings, retention of the invoice by the consumer can be made a precondition of an income tax deduction.

Generally speaking, the black economy is reduced by reducing the convergence of producer and consumer interests. A high rate of VAT encourages consumers to request that VAT is not invoiced. Income tax and social security contributions encourage vendors to suggest to their clients that they do not invoice them.

The problem with evaluating the impact of the measure is that it is difficult to measure activity which, by definition, is not directly observed.

3. The economic environment in the trial period

Altering the rate of VAT on certain labour-intensive services produces changes in the structure of the VAT base. Furthermore, depending on the VAT rates in force, the difference between the standard rate and the reduced rate varies among Member States.

Demand for labour-intensive services will not develop in the same way in a period of high growth or, at the other extreme, severe recession. How demand for the services concerned develops will be strongly affected by the point in the economic cycle at which the VAT reduction measure is applied. Therefore, before arriving at any conclusions about the impact of reducing VAT, one must take account of the economic environment.

3.1. Rates and scale of VAT

Applying a reduced rate of VAT to goods or services instead of the standard one can have different effects according to the levels of the reduced and standard rate. The knock-on effect on the consumer price which may be hoped for will therefore vary from one Member State to another. The table below shows the standard and reduced rates and the differences between them for each of the nine participating Member States.

The difference between the standard rate and the reduced rate of VAT ranges from 9 per cent in Luxembourg and Spain to 15 per cent in Belgium.

Table 1: VAT rates in force in 2000 (in %)

>TABLE>

Source: Commission departments

The table below shows each rate of VAT as a proportion of total transactions subject to VAT before the experiment with a reduced rate on labour intensive services.

The proportion of transactions subject to VAT charged at the standard rate is very varied and far from being 100%. It appears that many goods and services were already taxed at different, lower rates.

In six of the nine Member States participating the standard rate of VAT as a proportion of total transactions subject to VAT is lower than or equal to the EU average.

Table 2: Different rates of VAT as a proportion of total transactions subject to VAT in 1998 [5]

[5] Ranked by proportion at standard rate.

>TABLE>

Source: Commission departments

Reducing the VAT rate to below the standard rate for goods or services alters not only the structure of the VAT base, but also VAT revenue. The table below shows VAT revenue as a proportion of total taxation before the experiment with a reduced rate on labour intensive services.

Table 3: VAT revenue as a percentage of total taxation in 1998

Member State // VAT revenue as % of total taxation [6]

[6] Including social security contributions.

Belgium // 14.8

Greece // 21.0

Spain // 16.6

France // 17.6

Italy // 14.4

Luxembourg // 14.4

Netherlands // 17.1

Portugal // 22.0

United Kingdom // 18.4

EU // 16.6

Source: Commission departments

VAT revenue makes up an important proportion of total taxation and social security contributions. Among the nine Member States participating in the experiment, this proportion ranges from 14.4% in Italy and Luxembourg to 22% in Portugal. Six Member States are above the EU average.

3.2. General economic conditions

The measure applying a reduced rate of VAT to labour-intensive services officially came into effect on 1 January 2000. [7] Although the measure was not intended as a tool for the regulation of a particular economic situation, that situation is nevertheless a factor to be considered.

[7] Except in France, where it was applied from 15 September 1999 for renovation and repairs of private dwellings, and in Greece where it came into effect on 1 January 2001.

We need to know the state of the economy at the time in order to isolate the impact of the measure. The economic indicators below [8] give an overview of the economic situation and the major trends at the time.

[8] In the case of Luxembourg the business confidence indicator is used.

The economic confidence indicator is a survey, harmonised at Member State level, which takes account of both business and consumer sentiment. The indicator is the balance of positive and negative sentiment. It gives a picture of the economic cycle. The economic confidence indicator of the Member States participating in the experiment is shown in the graphs below with the EU aggregate indicator in bold.

Going by the graphs, the EU as a whole (line in bold) experienced growth from 1996 until approximately 2001, with a slight slow-down in 1998. The Member States' growth follows the same pattern, with slight variations during the slow-down phase in late 2000/early 2001. On the other hand, they were all in a growth phase when the measure was introduced.

>REFERENCE TO A GRAPHIC>

Source : Eurostat

>REFERENCE TO A GRAPHIC>

Source : Eurostat

>REFERENCE TO A GRAPHIC>

Source : Eurostat

This picture of the economic cycle is confirmed by the facts. The table below shows real GDP growth rates for the EU and the nine participating Member States from 1995.

The 1996 slow-down is reflected in the GDP growth rates, which in 1996 fell by 1.6% for the EU as a whole and also fell in most Member States. [9] Growth became dynamic again in 1997. The slow-down started in 2001.

[9] Of the nine Member States participating in the experiment, only Greece and Luxembourg had a higher rate of GDP growth in 1996 than in 1995.

In January 2000, when the VAT rate reduction on labour intensive services was introduced, growth was still dynamic. It did not begin to slow down until 2001.

Clearly then, the measure applying a reduced rate of VAT to labour-intensive services took place in the middle of a period of growth.

Table 4: Real GDP growth rates (in %)

>TABLE>

Source: Eurostat

The purpose of applying a reduced rate of VAT to labour-intensive services was to reduce unemployment. The table below shows unemployment trends in the European Union and the Member States participating in the experiment.

In the EU as a whole, unemployment has fallen constantly since 1998, with signs of bottoming out from 2002. Unemployment in the nine Member States also fell in line with this general trend. In some Member States this fall began one, and in some cases two, years earlier.

A reduced rate of VAT was, then, applied to labour-intensive services at a time when the fall in unemployment was well underway. In 1999, the year preceding the measure, the EU unemployment rate was at 8.7%, as against 10% from 1995 to 1997.

Table 5: Unemployment rates (in %)

>TABLE>

Source: Eurostat

Any reduction in VAT, if passed on in consumer prices, moderates price growth. The table below shows inflation trends from 1995.

The trend from 1995 to 2000 was for price growth to slow down. At EU level, inflation fell from 2.8% in 1995 to 1.2% in 1999. From 2000 inflation began to rise slightly again, at 2.1% for the EU as a whole. Inflation in the nine Member States more or less follows this pattern.

The VAT reduction measure was applied in the year when inflation began rising again in the Member States concerned. [10]

[10] In the case of the United Kingdom, the experiment only concerned the Isle of Man.

Table 6: Inflation, harmonised consumer price index (annual growth rate in %)

>TABLE>

Source: Eurostat

In conclusion, the experiment with a reduced rate of VAT on labour-intensive services took place at a time when growth had been dynamic for several years and there was a downward trend in unemployment and inflation. The overheating of the economy made itself felt in 2000 when the rate of inflation started rising again.

3.3. Economic indicators for specific sectors

The sectors chosen for the experiment with a reduced rate of VAT are of very little, if not negligible, significance in the economy as a whole, with one exception: that of renovation and repair of private dwellings. Using the information available at EU level, the economic situation of these sectors is described below.

3.3.1. Renovation and repair of private dwellings

Seven Member States chose this sector. Most Member States made the reduction conditional upon the age of dwellings, or restricted the types of works eligible. These conditions and restrictions are set out in the relevant sections of the Member State's assessment reports. In the case of the United Kingdom, this sector is geographically limited to the Isle of Man. [11]

[11] The United Kingdom is not, therefore, included in the discussion of the economic situation of this sector.

Renovation and repair of private dwellings falls within the construction sector, but statistics do not treat it as a clearly delimited sub-sector.

The confidence indicator in the construction sector is a survey constructed on the same bases as the confidence indicator presented above. However, this indicator only takes account of business sentiment in this sector. The graphs below show this indicator for the European Union as a whole (line in bold) and for the Member States which selected renovation and repair of private dwellings as one of the sectors for the VAT reduction experiment.

The confidence indicator for the construction sector in the EU as a whole follows the pattern observed in the analysis of the economic environment above, with the single exception being that there was no set-back in 1998. The sector thus went through a long period of expansion after the 1996 trough. Expansion slowed down in 2001 and continued in 2002. The overall pattern for the Member States is similar. The Spanish trend is more erratic, with the slow-down appearing earlier. Italy and Portugal did not go through a trough in 1996. France, on the other hand, enjoyed strong and uninterrupted growth from 1997 early 2001.

>REFERENCE TO A GRAPHIC>

Source: Eurostat

>REFERENCE TO A GRAPHIC>

Source : Eurostat

>REFERENCE TO A GRAPHIC>

Source : Eurostat

This reading of trends in the construction sector is reflected in the production growth rates shown in the table below.

Table 7: Growth: buildings production indicator (100 = 1995) Growth rate in %

>TABLE>

Source: Eurostat

At EU level, the growth of employment in this sector is in keeping with the growth of production. Among those Member States for which data are available, Belgium and Spain show parallel growth of construction and employment.

In France, on the other hand, the growth of employment is a year behind the growth of production. In other words the growth of production in the construction sector in France started up again in 1997, but employment growth did not do so until a year later. This lag is a result of labour supply constraints in the sector. According to the Institut National de la Statistique et des Etudes Economiques (INSEE), [12] recovery in this sector took place in 1998. This recovery grew much stronger from 1999, but there is a shortage of qualified labour. More than 30% of businesses stated they had difficulties recruiting sufficiently qualified labour. [13]

[12] INSEE 1 July 1999.

[13] INSEE 1 June 2000.

Table 8: Growth: employment indicator (100 = 1995) Growth rate in %

>TABLE>

Source: Eurostat

3.3.2. Other sectors

The other sectors have very little weight in the economy. Harmonised information at EU level is not, therefore, available. For instance, in the calculation of the harmonised consumer price index, the category of hairdressing and personal care services accounts for approximately 1%. The sector selected for the reduced VAT experiment only included hairdressing services.

Nevertheless, the harmonised consumer price indices allow a number of comparisons to be made.

By way of illustration, the table below shows, for the Member States concerned, the development of consumer prices in the index for all products and in the index for hairdressing and personal care salons.

In the year when the reduced VAT measure was introduced, inflation in the hairdressing and personal care sector was slowing down in Luxembourg and the Netherlands. In Spain, this slow-down remained very slight. The following year, on the other hand, inflation for these services rose sharply in the three Member States concerned.

It would therefore seem that reducing VAT only had a temporary effect on prices for the services concerned.

Table 9: Harmonised consumer price index (annual growth rate in %)

>TABLE>

Source: Eurostat

The next table shows, for the Member States concerned, the development of consumer prices in the index for all products and in the index for home repair and maintenance.

Again we see that, for home repair and maintenance too, prices dropped temporarily when the measure came into force, but rose sharply the following year in two (B & F) of the three Member States concerned.

Table 10: Harmonised consumer price index (annual growth rate in %)

>TABLE>

Source: Eurostat

4. Evaluation of the experiment

4.1. Evaluation method

The Commission's evaluation report [14] looks in detail at every stage of the economic mechanism for all the services concerned, using a common frame of reference, corresponding to the evaluation system discussed at the preparatory meeting in July 2002.

[14] SEC(2003) 622: Commission evaluation report on the "Experimental application of a reduced rate of VAT to certain labour-intensive services".

Evaluating the impact of reduced rates of VAT can be approached using a variety of methods. These may involve the analysis of economic data from national accounts or VAT declarations. Surveys may also be used directly focusing on the services concerned, or sectoral models (possibly general ones) may be used in combination with statistical or econometric tests. There is a broad range of possible methods, but the method selected should be the one which produces the most relevant answers.

Whatever method is selected, it is essential that the analysis shows the evolution over time of the sector concerned and of economic policy measures. Without knowledge of the sector prior to the measure's entry into force, it would be impossible to assess its impact.

In practice, the Member States applied the methods available in a variety of ways. The diversity of methods used, and of sectors to which the experiment was applied, makes it difficult to collate the assessments. To ensure maximum rigour and transparency, the examination of the Member States' evaluations follows the method of analysis approved in July 2002.

The answers provided by the reports to the points raised above fall into three categories. In some cases the facts presented lead to unambiguous conclusions. In some reports certain points are not considered for clear reasons, for example lack of data. In other cases, the information available in the report may not support firm conclusions.

The following paragraphs summarise the main results observed following this analytical framework.

4.2. Lower rates of VAT not fully reflected in prices

The first, essential step is for the reduction in rates of VAT to be passed on in consumer prices. If this does not happen at all, any increase in demand that might be observed would necessarily be the result of factors other than the VAT measure.

When they conducted price surveys, Member States found [15] that reduced rates of VAT were never fully reflected in consumer prices. Part of the VAT reduction is used to increase the margins of service providers.

[15] See Annex, Commission evaluation report on the "Experimental application of a reduced rate of VAT to certain labour-intensive services" SEC(2003) 622.

In Luxembourg, for instance, 40% of hairdressing services and 80% of shoe-repair businesses did not pass on the VAT reduction in their prices. One possible explanation may lie in vendors' aversion to risk. They prefer a risk-free approach (maintaining their prices at the same level and increasing their profits) to an approach involving risk (lowering their prices and hoping that increased demand will then increase their profits).

Renovation and repair of private dwellings seems to be the sector in which service providers most readily pass on the VAT reduction. This may be explained by the average scale of expenditure involved.

It has been found that where sectors -- such as shoe, bicycle and clothing repair in the Netherlands and repairs of dwellings in France -- pass on part of the reduction in VAT immediately, they tend subsequently to increase their consumer prices at a rate higher than inflation. Thus the application of a reduced rate only temporarily curbs the customary price increases.

In certain cases, most notably that of bicycle repair, a number of service providers refused to apply the reduced rate, arguing that the measure was too complex.

The difficulties encountered with passing on the VAT reduction in consumer prices show that this measure does not work on the lines envisaged by the Directive.

4.3. Expansion promoting demand

The economic environment, which was extremely favourable at the time when the measure was introduced, boosted demand and the turnover of providers of all services. The increase in demand observed can be explained by the increase in households' disposable income. In addition, other measures, particularly tax measures, were introduced and contributed to the growth of demand independently of the VAT rate reduction measure.

The Member States' reports do not allow the effects of growth, particularly vigorous at the time, to be distinguished from the possible effects of VAT rate cuts being passed on in part in prices.

4.4. No substantial impact on employment

Some Member States applied a pro-active policy to employment and assistance, concentrating on certain sectors, particularly construction.

In France, for instance, a set of measures to promote the acquisition of property and the restoration of housing had been introduced before the reduction in VAT rates. There was extraordinary growth in the sector at the beginning of 1999. Italy, as well as applying a reduced rate of VAT, authorised tax deduction of the costs of renovating and repairing private dwellings. Spain created a favourable tax environment for small businesses.

When the sectors tried to produce more to meet increased demand, they encountered difficulties with recruitment, or resorted to overtime. Portugal and the United Kingdom recognise that the measure was introduced at a time of full employment.

Under these circumstances, none of the reports of the Member States that participated in the experiment provide any solid evidence of the VAT reduction measure having an impact on employment.

Only France and Italy attribute the creation of a relatively large number of jobs (40 000 and 65 000 respectively) in the home renovation and repair sector to the reduction in VAT rates.

However, France's calculations do not take account of the 1999 storm, or of the tax measures promoting the construction of new housing and, to a lesser extent, the renovation and repair of housing. In addition, the construction sector in France had difficulties recruiting at the end of 1999. The figures presented in the French report also indicate that the cost of such job creation is very high, at about EUR 89 000 per job per year. Italy's report is based on a study of the professional organisation representing the sector. As in the case of France, the method of calculation used does not appear to take account of major direct tax measures.

4.5. A measure with purely relative impact

Empirical studies of the effects of reducing VAT rates generally in comparison with other measures show that reducing VAT is never the most effective way of stimulating job creation and GDP growth, and its budgetary cost is high in relation to its impact on the economy.

Specifically for this evaluation report, the Commission used its QUEST macro-economic model to carry out a simulation of a reduced rate of VAT compared with reduced labour charges.

Reducing VAT rates is always more costly in budget terms than other measures that directly target labour costs. These results are entirely consistent with the conclusions of previous studies.

Calculated at EU level, for the same budgetary cost a reduction in charges on labour creates 52% more jobs than a reduction in VAT fully passed on in prices.

4.6. Black economy: evaluation remains difficult

The Member States have provided little information on the impact on the black economy of reducing VAT. It was not possible to demonstrate that the former was reduced by the latter. Most of the results are no more than hypotheses unsupported by any demonstrated causal link between the measure and a reduction in the black economy.

On the other hand, independently of the VAT reduction measure, an important alternative way of combating undeclared labour was shown up: where the consumer can obtain an income tax deduction and must produce an invoice to do so, this appears to reduce recourse to the black economy. This effect was observed in buildings-related works.

5. General conclusion

In the experiment with reduced rates of VAT for labour intensive services, taking into account both information supplied by Member States and the inherent limitation of the analytical methods applied, it was not possible to find solid evidence of such reductions either boosting job creation. For the effect on the reduction on black economy the same observation can be made.

Price surveys carried out by the Member States indicated that the lower rate of VAT is passed on in consumer prices only partially or not at all. Furthermore, its impact on prices is generally only slight and does not appear to last. The mechanism of substantial price cuts leading to increased demand does not therefore work as envisaged by the Directive.

These results are also in line with the conclusions of previous studies. Compared with measures that directly target labour costs, the budgetary cost per job created by VAT cuts is always higher.

Specifically for this evaluation report, the Commission used its QUEST macro-economic model to carry out a simulation of a reduced rate of VAT compared with reduced labour charges. The assumption most favourable to the VAT reduction measure was made, namely that the full VAT rate cut is passed on in prices to the domestic consumer. Despite this favourable hypothesis, for the same budgetary cost a reduction in charges on labour creates 52% more jobs, calculated for the EU as a whole.

In view of all these results, and particularly given the difficulty of getting reductions in VAT rates passed on in prices, VAT does not, therefore, seem to be a good instrument for acting on demand through an impact on prices.

In terms of the internal market, given the services defined in Annex K to the 6th VAT Directive, there is little risk of distortions of competition. These Annex K services are essentially for purely local final consumers. There are only five categories of services in Annex K and Member States were not allowed to select more than three. Services differently defined in Annex K would probably have led to a risk of distortions. The Commission has received no complaints about any distortions of competition resulting from the application of reduced rates of VAT to these services.

It is now necessary to study future treatment of VAT for this type of services.

The issue will be studied as part of the biannual review of the scope of reduced rates provided for in Article 12(4). The review will be guided by the objectives of the new VAT strategy, [16] namely to modernise the common system of VAT, simplify it, strengthen it and apply it in a more uniform fashion.

[16] COM(2000) 348 final of 7 June 2000.

A proposal for a Directive will be drafted by the Commission and presented to the Council and the European Parliament in the course of 2003. In completing this study, the Commission will take account of the conclusions of this report and all the factors cited in its previous report on reduced rates of VAT [17] as well as all other relevant elements.

[17] COM(2001) 599 final of 22 October 2001.

The chief objective of the proposal will be to improve the working of the internal market by rationalising Member States' use of the reduced rates by giving Member States equal rights to apply them.

The characteristics of VAT should be borne in mind, most importantly the obligation to guarantee its neutrality, the need to define clear criteria for the application of reduced rates, the need to rationalise derogations and the long-term objective of more harmonised rates.