52007SC0910

[pic] | COMMISSION OF THE EUROPEAN COMMUNITIES |

Brussels, 5.7.2007

SEC(2007)910

COMMISSION STAFF WORKING DOCUMENT Accompanying document to the: Proposal for a COUNCIL DIRECTIVE amending Directive 2006/112/EEC with regard to certain temporary provisions concerning rates of value added tax and COMMUNICATION FROM THE COMMISSION TO THE COUNCIL AND THE EUROPEAN PARLIAMENT on VAT rates other than standard VAT rates

Main conclusions of the study on reduced VAT rates applied to goods and services in the Member States of the European Union {COM(2007)380} {COM(2007)381}

TABLE OF CONTENTS

1. Introduction 3

2. General observations 3

3. Member State interest: effect of reduced vat rates on economic growth and employment 4

3.1. Pass-through of VAT reduction – sector impact 4

3.2. Macro economic performance: promoting efficiency 5

4. Member State interest: the micro economic goal of promoting of equity 7

4.1. More equal income distribution 7

4.2. Promotion of merit goods/services 8

5. Community interest: effect of reduced rates on the functioning of the internal market 8

6. Conclusion 9

ANNEX: Copenhagen Economics' executive summary of their study on reduced VAT applied to goods and services in the Member States of the European Union 10

ANNEX List of VAT rates applied in the Member states – 01.05.07 13

1. INTRODUCTION

In section 2 of the Communication from the Commission to the Council and the European Parliament on VAT rates other than standard VAT rates[1] a short summary of the main conclusions of the study on reduced VAT rates applied to goods and services in the Member States of the European Union is presented. This study was carried out by Copenhagen Economics.

A summary of the study's results is further elaborated on in this working document. The Annex to this document contains the full executive summary of Copenhagen Economics' study.

2. GENERAL OBSERVATIONS

- Currently there are large differences in VAT rates both between sectors and within sectors in Member States and between Member States. About one third of the total EU 25 consumption is subject to non standard rates with the EU average non standard rate being 9 percent. Standard rates applied to the other two thirds of total consumption have an EU average of 19 percent.

- Reduced rates affect consumers' buying decisions and in this sense can be said to be distortive. The ex ante analysis of VAT policy scenarios show that a move towards more uniform rates would improve consumer welfare compared with the current situation. However, the changes in welfare would be relatively modest. Generally speaking, this holds in all policy scenarios in which the current non-standard rate structure is made more uniform or simplified (by adjusting the standard rate to ensure budget neutrality). It appears that in moving to a more uniform rate structure consumer welfare would (roughly) increase by 0,03 percent of private consumption compared to the current situation, i.e. by about 1,3 billion EUR. The smaller the variation in rates within a VAT system, the larger is the consumer welfare gain. In addition, by having a less complicated rates structure there are substantial savings in compliance costs for business and tax administrations (estimated at the double of welfare gains). It is clear that a simplification of the rates structure, especially within a particular sector, could contribute to reducing administrative burdens and improving regulatory quality Other VAT policy scenarios leave compliance burdens unchanged or increased.

- It should also be taken into consideration that other alternative policy tools , such as direct subsidies to particular activities, might achieve the same goals with fewer costs . Moreover, not only the potential, but also the effects of other policies already in place with the same aim need to be taken into account when introducing reduced VAT rates. It does not make sense to pursue the same policy objective with two or more different tools – this will only result in deadweight losses.

- If the aim of reduced rates is to induce a shifting of demand between sectors, mechanical revenue losses , in the sense of budgetary shortfalls arising whenever there is an initial high level of activity in the sector that is going to benefit from reduced VAT-rates but this is not matched by meeting the corresponding policy objectives, then this should be avoided. Mechanical revenue losses have a somewhat different character if the aim is to favour localised consumption by a certain part of the population, for the purpose of improving income distribution. In this case, mechanical losses arise from the degree of consumption in the other parts of population.

- An important rule appears to be that the more targeted the sector subject to a reduced rate is the more effective the measure will be. Here reduced rates will have less impact on other sectors and reduce (mechanical) losses in tax revenue and welfare. On the other hand, better targeting will result in higher compliance costs.

- Reduced rates are most effective where the goods/services are supplied directly to private consumers . Moreover, from an internal market perspective, for businesses there are very few cases where differences in VAT rates across countries can be exploited to reduce production costs and so the focus is naturally on consumers.

- Effective reduced rate measures need to be definitive or at least need to be perceived as such: producers are unlikely to respond strongly to VAT rate changes if they are perceived to be only temporary.

- The study suggests that the mix of instruments used to balance public budgets , e.g. following a reduction of VAT rates on certain products, should logically reflect the aims of the key policy objectives and the particular motivation to move in a particular policy direction. If, for example, VAT rates applicable to locally supplied labour intensive services are decreased, compensating this revenue loss by way of higher income taxes makes little sense as the whole rationale is to reduce the adverse effect from high taxes on labour.

3. MEMBER STATE INTEREST: EFFECT OF REDUCED VAT RATES ON ECONOMIC GROWTH AND EMPLOYMENT

3.1. Pass-through of VAT reduction – sector impact

- There is evidence of a (more or less) full price pass through of reduced VAT rates on a particular good or service in the final consumer prices , in turn leading to increases in consumer demand for this particular good or service and increases in sector production and employment. However, the strength and the swiftness of the response will depend significantly on a number of specific industry characteristics (e.g. consumer response to lower prices, the competition level within the sector, labour intensity and skills level, the level of black economy and do-it-yourself activities, etc.).

- If consumers react strongly to new lower prices , production and employment may increase significantly. Conversely, if consumers react only weakly to lower prices production and employment will not increase significantly. The type of the good will normally be quite decisive for the price response : for necessities (e.g. food) response to price changes will be much smaller than for more luxury purchases (e.g. package holidays, electronic equipment, household services).

- If production is labour-intensive there seems to be a stronger production and employment response to lower VAT rates in the industries affected. In sectors with limited competition pass-through to prices may be less than full, and thus the impact on production and employment may be weaker.

- If the sector is characterised by technology that is capital intensive or if the labour market is very tight or rigid , it may take some time before firms are able to adjust to the new market situation.

- Conversely higher VAT rates also have sizeable - negative - effects on demand and employment. Studies show that higher VAT rates have a stronger and faster response on prices than lower VAT rates. Consequently, when the reduction of VAT on some goods or services is compensated by higher VAT rates on some other goods or services -assuring budget neutrality- the effect of the rate adjustment of the reduced rate must outweigh that of the increased rates. Ultimate overall (long term) effects on total employment should also be considered. A necessary condition for lower VAT rates to be economically beneficial will be that there will be significant and important differences between the favoured and disfavoured products . These product differences apart, the only results will be distorted buying decisions leading to lower welfare within the Member States and in the Community, and higher compliance costs, without any economic enhancement, neither in job creation, economic growth or the functioning of the internal market.

3.2. Macro economic performance: promoting efficiency

- If the said significant differences are present, shifting activities between sectors might induce permanent changes in the economic structure of a Member State . Possible arguments supporting lower VAT rates on selectively chosen goods or services on the basis that it would increase overall economic efficiency are an increase in overall productivity and a reduction of structural unemployment.

- Lower VAT rates can increase overall productivity and GDP in a Member State if they can induce consumers to spend less hours on do-it-yourself (DIY) activities and more hours on their ordinary job and thus shift activities from DIY to the formal economy . A trained worker typically is more productive in his job than doing DIY. In addition tax revenue is likely to increase. Locally supplied services are industries where households have considerable scope for carrying out DIY (or buying from the black economy). There is a sound economic argument for (substantially) lowering VAT rates on locally supplied services (and restaurants) as the shift from DIY to formal activity is likely to lead to significantly higher overall productivity and a higher tax revenue. The argument is significantly stronger in Member States with higher marginal income taxes and larger initial VAT rates . Moreover, a further differentiation needs to be made: the benefits are increased for lower VAT rates on low-skilled locally supplied services , such as washing and cleaning than for high-skilled locally supplied services, such as housing repair. DIY activities play a relatively smaller role in high-skilled locally supplied services. However, other sectors, with equally high low-skilled workers are less relevant as they are not obvious candidates for DIY work, for example hotels or street cleaning. The productivity argument is, in particular strong for activities with limited need for formal training and specialised machinery . The same productivity argument holds, albeit to a lesser degree, if lower VAT rates can induce consumers to spend less money in the black economy . Similar gains in tax revenue can be expected.

- A number of countries already have national instruments in place to reap the said productivity benefits but the calculations in the study do not factor in the existence of these schemes. Calculations can thus not be used to advocate additional measures, especially because marginal benefits appear to fall when introducing more measures. Effects and the potential of other policies with the same aim need to be taken into account .

- According to simulations of a policy scenario, an extension of reduced rates to locally supplied services (hotels, restaurants and maintenance and repair related to social housing being included) to all Member States currently using reduced rates whilst keeping other reduced rates unchanged and raising the standard rate (to maintain budget neutrality), and taking into account the shift of activities between informal and formal economic sectors, (overall EU) measured GDP is suggested to increase by just over 1 percent (in a uniform rate scenario the GDP increase is estimated at 0,3 percent in a similar setting). GDP increases because some activities previously in the (non-measured) non-formal (DIY and black activities) sector are now being carried out in the (measured) formal economy. The increase in GDP overstates the real change as the activities are measured as though they were not carried out previously. When this is taken into account overall productivity is suggested to increase by roughly three quarters of a percentage of GDP .

- It is expected that lower VAT rates can reduce structural unemployment among low-skilled workers in a Member State if they can induce consumers to shift their demand towards sectors employing more low-skilled workers (hotels, restaurants, locally supplied services) than in the rest of the economy. In the process this would then boost their wages and increase incentives for employment on both sides of a rigid (low-skilled) labour market (in terms of high replacement rate, high unemployment benefits and high minimum wages). However, estimations made in the framework of the VAT policy scenarios tested, show that changes in demand only induce a miniscule increase in the demand for low-skilled labour in the scenarios where reduced rates are extended (amongst other to locally supplied services) and to a scale similar to the increase in demand for low-skilled labour arising from the uniform rate scenario. Consequently, de facto overall employment impacts are practically non-existent. Indeed VAT rates can only target the private consumption part of the sector’s output and sectors primarily using low-skilled workers are few and of limited economic importance. But, it is the case that there are significant differences in low-skilled employment within locally supplied services ; cleaning services in private households appear to have a much higher share of low-skilled employment than renovation and repair of private dwellings, hairdressing and domestic care. There may thus be a better argument for targeted reductions in VAT rates on carefully selected parts of locally supplied services rather than for general reductions for all of them. Moreover, reduced rates are most effective where the goods or services are supplied directly to private consumers . Cleaning in private households, domestic care and hairdressing appear to be the best sectors in this respect.

- For addressing structural employment problems of low-skilled workers alternative instruments , such as more general labour market policies may have more potential: easing of employment protection legislation and educational policies raising educational attainment levels. One central question is whether the position of low-skilled workers and the welfare of society as a whole are better off by introducing new measures to affect labour supply or instead to expand the demand for this group of workers.

4. MEMBER STATE INTEREST: THE MICRO ECONOMIC GOAL OF PROMOTING OF EQUITY

- In terms of other policy objectives, it is argued that carefully designed reductions in VAT rates may solve concerns mostly related to the question of equity . More particularly it is argued that reduced VAT rates may generate a more equal income distribution (after consumption ) if they make cheaper goods and services that are mostly consumed by low-income groups or promote consumption of (merit) goods or services that governments deem to have intrinsic society values in excess of the consumption value for the individual consumer at normal VAT rates (e.g. cultural goods, such as books and tickets for museums and theatres).

4.1. More equal income distribution

- Consumption shares need to differ sufficiently between low and high-income groups and need to be stable over time for reduced rates to be an effective tool in terms of income redistribution. Consumption patterns across households tend to be relatively equal in countries where the overall distribution of incomes is relatively even. The higher initial equality, the less potent are reduced rates for promoting further income equality. On average the largest difference in consumption between the lowest and highest income group appears to relate to food and to utilities like electricity and heating . However, huge differences exist across the Member States .

- Alternative tools to reducing VAT rates might be very relevant, in particular in Member States with initially even income distributions. For a Member State with a well developed and comprehensive social system including other schemes of targeted support, direct transfers to families in need are likely to be more cost-effective with respect to compliance costs, mechanical revenue losses and adverse effects on consumer welfare. Another alternative could be labour market reforms by improving employment possibilities for low-income households. Progressive income tax scales are also a powerful means to reduce after-tax income inequality.

- As regards the ex ante analysis of VAT policy scenarios, it appears that the uniform rate scenario is regressive making the income distribution more skewed (it increases prices for low income households and lowers prices for high income households). The scenario where reduced VAT rates are applied to locally supplied services or where reduced rates are extended leaves the income distribution largely unaffected. The scenario where rate derogations are abolished is slightly regressive (reduced rate variations are compensated each time by a modification of the standard rate).

4.2. Promotion of merit goods/services

- Member States may be able to increase consumption of merit goods, but they should seriously consider the economics of such moves in particular in an environment where other policy tools are in place. Once again compliance costs may be significant and issues of fiscal neutrality between like products and of distortion of the internal market may be raised. Moreover, the policy goals, i.e. a more even income distribution and promotion of merit goods or services, may be in internal conflict, in certain cases.

5. COMMUNITY INTEREST: EFFECT OF REDUCED RATES ON THE FUNCTIONING OF THE INTERNAL MARKET

- AS FOR VAT DIFFERENCES WITHIN A MEMBER STATE, VAT DIFFERENCES BETWEEN MEMBER STATES HAVE THE POTENTIAL TO DISTORT consumers' buying decisions, to cause an efficiency loss and to change the pattern of trade between Member States and thus to redistribute VAT revenue between Member States. For these reasons, the VAT structure has clear implications on the functioning of the internal market.

- The areas where significant differences in rates across Member States risk distorting competition in the internal market are cross border sales where the final consumer physically moves to another Member State to purchase a good or service, consuming it on-the-spot (e.g. tourism, restaurants,) or bringing it to his own Member State (e.g. clothes, sports and electronic equipment, music CD, movie DVD) and distance sales, including e-commerce which is increasing and, where a final consumer orders a good or service in his own Member State from another Member State, subsequently having it brought to his own Member State (e.g. e-trade in (e-) books, music…. ).

- The main characteristics of goods being particularly suitable for cross-border shopping are high price per unit weight/volume, low perishability (during transport) and with little cultural specificity.

- The same characteristics as mentioned above are equally likely to be important for distance sales where a high level of product (brand) credibility is to be added, because the consumer is not able to see and try the product in advance.

- There do not seem to be major reasons to be worried about the consequences for the internal market in respect of many goods and services. It is difficult to see any concern related to differences in VAT rates on purely l ocally supplied services . Those sectors where concern may be well-motivated are books and medicine . Services provided by restaurants and hotels are somewhere in between, because they are mainly directed at domestic consumption, but may also affect distribution of tourism between Member States and may have a non-negligible impact in border regions. Possible distortion as regards restaurant and hotel services is likely to be different in magnitude across Member States (stronger for smaller and/or tourist oriented areas) and the degree of possible substitution of holiday destinations plays an important role. As far as the business consumption of these services is concerned, the rules on VAT deductibility may also impact on the functioning of the internal market. Culture and entertainment in the context of tourism and educational services may also be an area of concern. Generally, the standard rate also becomes increasingly important given the significantly growing importance of e-commerce, distant sales and (digital) services.

- Calculations done for the design of different VAT policy scenarios (in a General Equilibrium Model) show that some of the same sectors that are likely to have relatively significant problems with widely differing VAT rates, e.g. books/medicines and culture/entertainment are among those sectors with the highest variability of VAT rates between Member States. The scenario that drives down this variability between Member States is the uniform rate scenario, rather than the scenario which is explicitly designed to level out the reduced rates (by extending them) between Member States.

6. CONCLUSION

The current VAT structure with lower rates for some goods and services leads to extra costs for the Community and Member States. For this reason, Member States contemplating lower VAT rates should carefully consider the expected size and likelihood of benefits they hope to achieve by this move. And the Community should carefully examine the consequences for the proper functioning of the internal market .

Moreover, using the VAT system to attain one policy objective might be at the cost of other goals. The overall VAT configuration that in conjunction with other policy instruments best meets overall objectives needs to be established (e.g. is the VAT reduction an additional measure to another measure intended to achieve the same goal, do the expected benefits of a VAT reduction in a particular sector outweigh possible disadvantages for other sectors, the knock-on effects of assuring budget neutrality, and extra compliance costs). Furthermore, VAT rate reductions are not necessarily the best tool to fulfil the desired policy goal. For this reason, it is always important to seriously consider alternative policy tools.

However, for some very particular sectors, where reduced rates are currently either applied or not, there seems to be rather compelling economic arguments for maintaining, introducing or even extending reduced VAT rates . This does not only rely on the characteristics of the sector but also on the particular economic environment of the Member State. However, the resulting higher compliance costs need to be taken into consideration when such decisions are made. For these sectors there appear to be no significant problems for the functioning of the internal market. For other sectors it is hard to find convincing economic arguments, either based on efficiency or equity, for maintaining reduced VAT rates. Instead there is an economic rationale for making rates more uniform in order to alleviate the different distortions arising from differentiated VAT rates.

ANNEX: Copenhagen Economics' e xecutive summary of their study on reduced VAT applied to goods and services in the Member States of the European Union

Value Added Tax (VAT) in Europe is regularly subject to intensive debate. It is often argued that the current VAT system should be made more uniform to enhance economic efficiency and to protect the functioning of the internal market. But it is also regularly argued that extending reduced VAT to this or that particular product would create economic benefits such as more employment and less inequality. Which side has the upper hand?

This study argues that there is a strong general argument for having uniform VAT rates in the European Union. Uniform rates is a superior instrument to maintain a high degree of economic efficiency, to minimise otherwise substantial compliance costs and to smooth the functioning of the internal market. However, this study also argues that there are exceptions. There are real and valid economic arguments for extending lower VAT rates to some very specific sectors in member states characterised by specific economic structures.

This study examines the theoretical and empirical merits of four different arguments for reduced VAT rates. Two based on efficiency grounds: reduced VAT can increase efficiency by increasing productivity or by reducing structural unemployment. Another two based on equity grounds: reduced VAT can enhance equity by improving the income distribution or by making particular products more accessible to the entire population. The pros and cons of these arguments can be summarised as follows:

First, there is a convincing theoretical and empirical argument for extending reduced VAT rates (or other subsidies) to sectors whose services are easily substituted for do-it-yourself or underground work, e.g. locally supplied services and some parts of the hospitality sector.

The argument is that high tax wedges (high marginal income tax and high VAT rates) make it very expensive to buy these services on the market and more attractive to do yourself. The implication is that high skill professionals spend time on low skill work at home instead of spending time with their families or increasing their more productive labour supply. Lower VAT rates serve to counter this development. Simulations indicate that the gains in welfare, productivity and GDP are sizeable in all member states, even though the largest gains by far will accrue to member states with high tax wedges. Reduced VAT rates are not expected to have negative implications for the functioning of the internal market as the relevant products are typically not traded across EU-borders.

Second, we claim that there is a theoretical but not an empirical argument for extending reduced VAT rates to sectors employing many low skill workers in order to boost low skill demand, e.g. hotels, restaurants and locally supplied services. However, there may be a case for a limited, supplementary role via carefully targeted reductions in the context of grander labour market reform.

The theoretical argument is that reduced VAT rates, by boosting demand for such services, stimulate demand for low skill workers, and push up their wages such that employment becomes a more attractive option than unemployment. The argument only holds in member states with rigid and non-flexible labour markets for low skill workers. In other member states the argument is not appealing, as increased demand may just stimulate wages for this segment of the labour market.

However, simulations indicate that the overall impact on demand for low skill workers is unimpressive because differences in low skill employment between industries are limited. Indeed, making standard VAT rates apply for all sectors currently benefiting from reduced rates is likely to create a similar sized demand boost for low skilled workers. If implemented, reduced VAT rates may have some limited implications for the functioning of the internal market, in particular through tourism.

Third, we claim that there is a limited and contingent argument for extending reduced VAT rates (or other subsidies) to sectors particularly favoured by low income households in order to improve the (post consumption) income distribution, but the argument only holds for member states with significant and stable consumption differences between high and low income groups. In reality, the only relevant sector is food.

The argument is simple Reducing VAT rates on food which constitutes a larger share of consumption for low income households than for high income households implies a cost saving that is particularly beneficial for low income households. The larger the difference in consumption shares is, the more effective the argument becomes. Simulations show that the argument has some empirical support in member states, in particular those with high initial income inequality.

However, extending reduced VAT rates to food also brings about significant complications. Compliance costs seem to be particular large for the food sector due to its multitude of products, the enticing inclination to treat healthy and non-healthy food differently and the existence of a grey zone between sale of food in supermarkets and prepared food in restaurants. In addition, reduced rates on food tend to make do-it-yourself-cooking economically more attractive relative to frequenting restaurants. For this reason, they tend to compound restaurants’ disadvantage competing with home made meals even when VAT rates are similar and to offset any productivity gains to be gained from eventual reduced rates on restaurants, cf. above. Finally, it is not clear that the best instrument for improving the income redistribution is the VAT system, in particular not in member states with broad, well-developed social security systems. Reduced VAT rates on food are not likely to have significant consequences for the functioning of the internal market.

Fourth, we claim that there is a limited and contingent argument for extending reduced VAT rates (or other subsidies) to sectors that for some (good) reason are under-consumed. The motivation can be to make cultural (merit) goods more available for low income households or to stimulate consumption of goods with positive externalities. Examples of the former could be books, music and cultural events; of the latter energy saving appliances.

The argument is simple: demand can be boosted on any product by lowering VAT rates. However, it is often difficult to verify whether low income households in reality are induced to purchase more merit goods or whether the lower rates in reality serves as a subsidy to high income households initially consuming more merit goods. In addition, extending lower VAT rates to e.g. energy-saving appliances have limited effects on CO2–emissions if they are covered by other regulatory instruments such as emission trading schemes and may give rise to non-trivial complications for the functioning of the internal market. Furthermore, effects on total energy use are ambiguous: it may well switch consumption from less to more energy efficient hair dryers for example but may at the same time switch overall consumption in the direction of energy intensive products (more hairdryers). Reduced rates on some merit goods such as books and music tend to create some serious tensions with the functioning of the internal market, primarily due to the ease of electronic trade.

The study also identifies a number of concerns that should be carefully evaluated in each specific case:

- Most arguments in favour of lower VAT are equally valid for other policy instruments, for example targeted subsidy schemes or targeted changes in income tax. However, this study does not evaluate whether lower VAT is the best instrument within the group of feasible instruments. For this reason, in any specific case it should be evaluated carefully whether lower VAT is the best instrument to achieve the desired effects or not.

- All empirical evidence indicates that compliance costs associated with lower VAT rates can be sizeable. Differences in VAT rates between similar products may in particular give rise to a substantial number of administrative and legal conflicts about the proper classification of specific goods. Swedish estimates indicate that such cases have very significant costs for the society.

- The key to an efficient application of lower VAT or any other subsidy is to keep low mechanical revenue losses. Mechanical revenue losses arise when lower VAT (or any other subsidy) is ceded to consumption that does not contribute to reaching the desired goal. For example, if lower VAT is ceded to food in order to improve the income distribution there will be a mechanical revenue loss because high income households will also benefit from lower VAT.

- The choice of financing scheme to secure budget neutrality should be carefully considered in the context of the goals to be achieved by reduced VAT rates. For example, if lower VAT on locally supplied services (in order to increase productivity) is financed by higher marginal income taxes, the desired effect may be nullified or reversed. If lower VAT on food (in order to improve the income distribution) is financed by higher VAT on items primarily consumed by high income households, the desired effects may be reinforced.

To summarise, there seems to be a strong argument for making the current VAT structure more simple and uniform, but also an argument for selective cuts in VAT rates primarily in locally supplied services and parts of the hospitality sector. We stress that the current study cannot be seen as a per se endorsement for further reductions in VAT. The devil is in the detail and we stress the need to consider each case on its own merits and to seriously appraise whether alternative non-VAT instruments may be preferable to reduced VAT rates.

ANNEXList of VAT rates applied in the Member states – 01.05.07

[pic]

[1] COM(2007) 380 final