The value-added tax systems of almost all EU member states employ reduced tax rates in areas such as foodstuffs and pharmaceuticals. There is a historical background to this. Europe has a long tradition of general consumption taxes, and reduced tax rates for some goods were already in use in the consumption tax systems in existence before the VAT was introduced. The VAT was pioneered in France in the 1950s, and when it was adopted by other countries, they often carried over the existing reduced tax rates to the VAT. Social policy and redistributional concerns also played an important role in the introduction of reduced rates. However, in the EU system, reduced rates are optional, and indeed some countries make limited, or even almost no, use of them.

There are problems associated with reduced tax rates. For one thing, reduced VAT rates, together with exemptions, tend to push up the standard rate. In the EU, the cumulative effect of reduced rates and exemptions eats up around one-third of VAT revenue. European governments have relied on increasing the standard rate in order to compensate for this loss of tax revenue. This means that reducing the rate on some items has the effect of increasing the rate on all other items.

Another problem is that reduced rates complicate the tax system, increasing the cost of levying and complying with taxes. This problem, and particularly the need to define precisely the nature of the goods for tax purposes, is more serious than one may think at first sight. Multiple tax rates impose more control costs on the tax administration, which are financed by tax money. In addition, when the scope of the tax reductions changes or expands, the burden of accounting procedures (employment of VAT specialists, etc.) on companies and business owners increases, increasing compliance costs. It is difficult to reduce these costs even if an invoice system is used.

In part because of this, the EU has never, generally speaking, encouraged member states to broaden the application of reduced rates. The EU Directive on VAT stipulates that there should be no more than two reduced tax rates, and that the rate should be no less than 5%. In some cases, the European Commission has also recommended individual member states to limit the scope of reduced rates and exemptions. Nevertheless, some countries apply four or even five reduced rates, and some use rates of less than 5%. The EU has no choice but to allow these exceptions.

Once a reduced rate is granted, it is very difficult to take back. As a result, the scope of reduced tax rates tends to grow over time. It is generally quite easy to find some justification for cutting taxes on a particular product or sector, and the political pressure to broaden the application of reduced tax rates is often fairly intense.

New Zealand has had its goods and services tax (GST) system for nearly 30 years*1, and unlike western European nations, did not carry over multiple tax rates from its previous wholesale tax system when the system was introduced. The GST was applied evenly to all products, and there were minimal exemptions. Considering the example of European nations, which introduced reduced tax rates as a result of their specific political circumstances, but later found that exemptions continued to multiply, it was considered wisest not to introduce them from the beginning. The explicit policy in New Zealand was to efficiently ensure tax revenue via the broadest possible base and the lowest possible rate.

In New Zealand also, it was argued that reduced tax rates should be applied to specific product groups such as foodstuffs. But reduced tax rates are a very inefficient means of targeting regressivity. Is it really lower income earners who buy the majority of these products? Newspapers, for example, might be something that is purchased more by well-off classes. Reduced tax rates actually offer a benefit to people who do not really need that benefit.

Rather than reduced tax rates, New Zealand’s approach has been to respond by offering tax exemptions targeting low income earners, and providing social security benefits to support low income earners in their daily lives. Singapore, which has adopted a system very similar to the New Zealand system, has established a voucher system for low income earners, enabling them to receive GST refunds in relation to certain products.

Because New Zealand applies GST to almost all goods and services at the same rate, it is able to make use of the innate strengths of the system and efficiently obtain revenue without excessive administrative costs. The GST rate was maintained at 12.5% for 21 years following its introduction in 1989*2, and even today the level is far lower than the levels in effect in European nations, but New Zealand has maintained fiscal soundness.

The lack of reduced tax rates also makes it easy to predict long-term tax revenues in New Zealand’s system. When tax rates differ for different products, tax revenues fluctuate depending on which industry grows and which industry shrinks in the future. By doing away with reduced tax rates, tax revenue becomes stable, and that enables us to formulate policies with a view towards the future.

*1: 1 October, 2016 marked the 30th anniversary of the introduction of GST.
*2: Since 1 October 2010, New Zealand’s GST rate has been 15%.

Sweden is known for its well-developed social welfare system, and, in addition to income tax, the government relies on value-added tax for a significant proportion of its revenue. Swedish citizens accept a high 25% rate of value-added tax because tax income and government expenditure are directly linked, and they feel that they get good value for the money in the form of the benefits of social welfare policies.

For example, I did not have to save any money to send my children to university. If, hypothetically, the government reduced taxes, but stopped paying for university tuition, reduced pensions, and changed to a system in which citizens were forced to purchase private health insurance, citizens would protest.

Value-added tax applies the same tax rate to high-income earners and low-income earners. Because of this, some people argue that it is contrary to the principle of redistribution of income, and is not fair. However, there is an entirely different understanding in Sweden. Because the value-added tax is a fixed rate, the burden is higher for wealthier people who spend more. Besides, because the revenue from value-added tax funds expenditure on universal welfare measures, citizens can see that there is an effective redistributive function. The Swedes also have a prevalent ideology of fairness, and believe that even people who cannot afford it should have access to reasonable medical care and education.

In Sweden the rate of value-added tax has been reduced in certain sectors, for example restaurant services, and there has long been discussion of extending the reduction to food in general. However, this seems to be more the result of industry lobbying than consideration for low-income earners. There seems little good argument for encouraging people to eat more in an era in which food is abundant, or for extending the reductions to such foods as caviar and foie gras.

Progressive taxes are not necessarily considered fair. There was a higher degree of progression in income tax rates in Sweden in the past. High-income earners did not view the tax as legitimate, and did all they could to avoid paying tax.

The Social Democratic Party which created the welfare state placed an emphasis on fiscal soundness, and therefore used value-added tax as a source of revenue for the welfare system. Support for the welfare state among Swedish citizens has not wavered as a result of globalization, but has actually become stronger.

If asked whether they would prefer to receive 10,000 yen today or 10,100 yen in a week, most people will choose the former. However, if asked whether they would prefer to receive 10,000 yen in a year, or 10,100 yen in a year and one week, most people will choose the latter. Or again, many people will be quite sanguine regarding a choice that takes effect a year in the future, but when a year has passed, will wish to change the choice made by their selves of one year ago. In behavioral economics, this is termed “present bias.”

Initially scheduled for October 2015, the increase in Japan’s consumption tax rate was postponed for 18 months in November 2014, a year away from implementation. In June 2016, it was announced that the increase would be postponed for another 30 months, and would be introduced together with reduced tax rates which will have a limited redistribution effect and will distort consumption behavior. The increase was supported when it was envisioned for the distant future, but as the date for the introduction of the increase has grown nearer, it has now been postponed twice.

The Japanese public shares the awareness that a future increase in the consumption tax rate will be essential if the nation is to avoid fiscal collapse. The public therefore approves of the government’s commitment to increase the tax rate. However, as the scheduled date for the increase approaches, the public prioritizes present gain, and supports the government in backing down from its commitment. This is because the public is caught up in present bias. But it is the public itself which will have cause to regret these postponements.

“Loss aversion” is another concept supplied by behavioral economics which can help in explaining the postponement of the increase in the consumption tax rate. This refers to the tendency, when facing a loss, to choose a gamble involving the potential for an even greater loss rather than accepting the loss, if there is a possibility that the status quo might be maintained. In the context of the consumption tax increase, in which the choice is between accepting an increase in the tax rate and a consequent loss now, or a gamble in which the result is likely to be fiscal collapse and high inflation, but in which fiscal soundness might be realized through economic growth, this means choosing the latter.

In order to avoid postponement of the introduction of the increase in consumption tax due to present bias, some commitment device which is binding on government action is necessary. However, the breaking of a public pledge twice has resulted in a loss of trust in that commitment. Preventing the public from choosing a policy that represents a gamble due to loss aversion will necessitate framing of the issue in such a way that the tax increase tends not to be seen as representing a loss. Measures of this type based on behavioral economics must certainly have been put into effect in countries like New Zealand and Sweden, which have been successful in increasing taxes.

The stated purpose for the introduction of reduced tax rates for Japan’s consumption tax (value added tax) is to ease the regressivity, but, it actually has the effect of exacerbating the sense of income inequality because it is equivalent with a policy to distribute more money to high income earners. A simpler and more effective method is available to ease regressivity.

A numerical example will help. Let us consider a low income earner with a salary of 200,000 yen per month, and a high income earner with a salary of one million yen per month. The low income earner spends 150,000 yen per month, of which 50,000 yen represents expenditure on food; the high income earner spends 440,000 yen per month, of which 100,000 yen represents expenditure on food.

If a tax rate of 10% is applied to all consumption, the low income earner and the high income earner will pay 15,000 yen and 44,000 yen in tax respectively. In terms of the proportion of their respective incomes, the low income earner pays 7.5%, and the high income earner pays 4.4%. In this sense, consumption tax is regressive.

Let us now assume that a reduced tax rate of 8% is applied to food. The tax burden on the low income earner is now 14,000 yen (100,000 yen × 10% + 50,000 yen × 8%), and the tax burden on the high income earner is 42,000 yen (340,000 yen × 10% + 100,000 yen × 8%). The low income earner now pays 7.0% of their income, and the high income earner pays 4.2%. Thus, the reduced tax rate on food produces a slight easing of regressivity.

However, if we compare the absolute value of the reduction in tax, the reduced tax rate reduces the tax burden on the low income earner by 1,000 yen, while the tax burden on the high income earner is reduced by 2,000 yen. Of the 3,000 yen “saving” by the reduced tax rate, the low income earner receives one-third, and the high income earner receives two-thirds. While it is technically correct to say the regressivity of consumption tax has been eased, no-one would claim that the economic disparity has been reduced.

It is simple to achieve a greater easing of regressivity using the same 3,000 yen. This would involve distributing the 3,000 yen equally to each individual. This 1,500 yen return would make the tax burden on the low income earner 13,500 yen, and the tax burden on the high income earner 42,500 yen. Relative to their incomes, this represents 6.75% and 4.25% respectively.

Compared to the introduction of reduced tax rates, which are complex, involve high administrative costs, and produce a considerable distortion of economic activity, the equal distribution among the general public of an amount equivalent to the reduction in tax revenue resulting from the reduction in the tax rate would be both simpler and have a greater effect in easing regressivity.