Study on reduced VAT applied (2007)

Study on reduced VAT applied to goods and services in the Member States of the European Union - Final Report

Executive summary

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 a broader 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 so 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 and prepared food. 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:

1. 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.
2. 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.
3. 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.
4. 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.