Goods and Services Tax - Zero-Rating for Foodstuffs, Books, Newspapers and Magazines (P.169-2007) – comments

STATES OF JERSEY

r

GOODS AND SERVICES TAX: ZERO-RATING FOR FOODSTUFFS, BOOKS, NEWSPAPERS AND MAGAZINES (P.169/2007) – COMMENTS

Presented to the States on 19th November 2007 by the Minister for Treasury and Resources

STATES GREFFE

COMMENTS

Summary

Senator Shenton has put forward a proposal to zero-rate Goods and Services Tax (GST) on food, books, newspapers and magazines in order to protect Islanders on low-to-middle incomes.

Although the objectives behind his Proposition may be laudable, this Proposition is not a good way to achieve those aims.

I acknowledge and support the desire of the public and of States Members to use some of the recent improvements in tax receipts, arising from the success of our fiscal policies, to reduce the impact of GST on low- to-middle income households. Blanket GST exclusions are, however, a costly and ineffective way of doing this.

The States has now approved the Income Support scheme which will fully protect those on low incomes from GST on food, books, newspapers and magazines, so the proposed exclusions would reduce the cost of food for those households on Income Support, so they would benefit by a total of about £400,000 per annum.

Whilst providing little benefit to households receiving Income Support, the proposed zero-ratings would benefit those low-to-middle income households just above Income Support thresholds to the tune of £58 a year. They would, however, be of even greater benefit to high income households, who would gain up to £138 a year.

Indeed, of the £3.1 million in tax revenues lost from the proposed exclusions, approximately£1 million would go to low-to-middle income households, but £2 million would go to high income households. This is because the well-off spend far more on food, books and newspapers than those on low-to-middle incomes.

Accordingly, this amendment does not do what it intends to: it provides relatively little assistance to those on low incomes, who receive Income Support; it provides only a small amount of assistance to those on low-to-middle incomes; but makes the households with the highest incomes considerably better-off.

A far preferable way of sharing our improved tax revenues with Islanders on low-to-middle incomes would be by further increasing income tax thresholds. The cost of Senator Shenton's proposals is estimated at £3.9 million p.a. Instead of zero-rating food, books, newspapers and magazines, the money could be better used by maintaining the Income Support of £400,000 and spending the remaining £3.5 million increasing income tax exemptions. The thresholds could be increased by 6.5% instead of the 3% proposed in this year's budget. If this was done, households on Income Support will be at least as well-off as under Senator Shenton's Proposition, whilst other low-to-middle income households would be up to 3 times better off than they would from zero-rating food, books and newspapers.

The extra saving to a typical low-to-middle income household would be as much as £180, compared to the £58 they would save under Senator Shenton's proposed GST exclusions.

A further disadvantage of the proposed exclusions is that they would increase enormously the complexity of GST with significant costs for both the States and small businesses. The States would need to employ eight extra staff to administer these exclusions and incur approximately £800,000 in additional administrative costs.

Many small businesses with a turnover below the £300,000, which previously did not have to administer GST, would also now inevitably be forced to register for GST. This is because in order to minimise their competitive disadvantage compared to large retailers, particularly those with U.K. links (who will have the systems to process zero-ratings), small retailers will wish to recover the GST they suffer on their costs, and the only way they can do this is by registering and following the entire GST procedures.

In summary, the exclusions proposed by Senator Shenton, would lose £3.1 million in tax revenues, cost£800,000 for the States to administer, hit small businesses, yet only provide £1 million in support to households on low-to- middle incomes the Senator seeks to assist.

There is no doubt that our fiscal policies are working, and this has resulted in improved tax receipts. I also recognise the concerns of both the public and States Members about the impact of GST on low-to-middle income households, and the desire to use some of the improved tax receipts to benefit this group. Blanket GST exclusions are an ineffective, costly and administratively expensive way of helping these households.

If the States agrees with me, and rejects these zero-ratings, I will commit to bringing an amendment to this year's budget to increase income tax thresholds by 6.5%, rather than the 3% currently inscribed in the budget. I will also work with the Minister for Employment and Social Security and will also ensure that those who receive Income Support receive the same improvement as they would under Senator Shenton's Proposition.

Increasing income tax thresholds would give low-to-middle income households on average 3  times more money than  they  would  save  from  zero-rating  food,  newspapers,  books  and  magazines. It  would  also  incur  no administrative cost, keep GST simple, and avoid small businesses having to deal with complex and expensive administration.

The States is recommended to reject the zero-rating of food, books, newspapers and magazines and instead indicate its support to providing far greater assistance to low-to-middle income households by increasing tax thresholds, by a further 3.5%, to 6.5% and maintaining additional Income Support of £400,000.

The following table compares the effect of Senator Shenton's Proposition with my own proposal:

Table 1 Background

In 2004 the States agreed to introduce the 0/10% corporate tax structure in order to retain the finance industry and hence secure the economic future of the Island.

However, the overall effect of "zero/ten" will be to reduce Jersey's future annual tax revenues by approximately £100  million. The main impact of this will be felt in 2010 and the full effect by 2012.

In order to fill this anticipated revenue gap the States agreed a package of measures that included:

restrictions on States spending;

an economic growth plan;

an Income Tax Instalment System;

a phasing-out of certain Income Tax allowances for higher income groups ("20 means 20");

the introduction of a Goods and Services Tax (GST); and

a revised Income Support system.

The overall effect of this package, because of Income Support, plus the phasing-out of income tax allowances for those on higher incomes, is progressive, i.e. those on low incomes will pay no extra tax, and everyone else in society will pay more tax with those on high incomes paying most.

Previous Votes to Propose Exclusions

Projet 169/2007 is not the first attempt to exclude items from GST coverage. States Members will recall 3 previous attempts during the past 2½ years – one in May 2005 (P.44/2005 Amendment), one in September 2005 (P.165/2005) and the other in October 2006 (P.86/2006). All 3  proposals were overwhelmingly defeated, for good reasons.

Table 2 below gives the voting record of these previous debates:

Voting Record of Previous Attempts to Exclude Basic Items from GST

Table 2

Although none of the principles have changed since those debates, it is worth rehearsing the reasons why States Members decided so convincingly they wanted a low-rate, broad-based and simple GST.

Why Choose a Simple Goods and Services Tax (GST)?

A simple GST is one that has a broad-base and a single positive rate. It requires few zero-rates (other than for exports  and  international  transport  of  goods  and  persons,  the  supply  of  residential  accommodation);  few exemptions (beyond the usual ones for small traders, the financial sector, postal services, etc.); and an invoice- based collection and administration system, with as few special schemes as possible.

Being a tax with a single positive rate, the simple GST minimises the costs of compliance for the traders and suppliers. The costs of administration for the States are also low.

The simple GST also ensures that the effective burden of the tax on the consumer is exactly the same as the nominal rate of the tax and the customer knows exactly what he is being charged by way of GST. The tax therefore treats all consumers fairly.

A simple GST treats all businesses uniformly, with minimum deviations, and thus minimises the distortions in the allocation of resources in the economy. It also maximises the revenue yield for the States at the lowest possible tax rate.

A complicated GST, on the other hand, is one that consists of many more exclusions (mainly zero-ratings, exemptions and special schemes) all of which tend to narrow the tax base, complicate tax administration and make tax compliance cumbersome and costly.

Traders with a mixture of sales of zero-rated, exempt, and taxable supplies have to keep separate accounts for each of these categories of sales, imposing on them a significant additional burden of compliance. Such traders can also easily be tempted to evade taxes on their taxable supplies, but even if not attempting to evade, can innocently make errors on their (now complicated) returns which need investigating and correcting, thereby adding to administration costs.

Extensive zero-ratings and exemptions generate continuing pressures from taxable sectors for equity and therefore zero-ratings, exemptions, or special treatments for them as well.

By virtue of its narrower tax base, the complicated GST also requires a higher rate to yield a given amount of revenue than does a simple broad-based GST – i.e. fewer items attract a higher rate of tax to achieve the same revenue yield.

The traditional, complicated model of GST has, for the reasons above, generally been superseded throughout the world by the simple broad based model of GST, with few exclusions.

The GSTs of Singapore and New Zealand are examples of simple, modern GSTs while Value Added Tax (VAT) in the United Kingdom is an example of the traditional complicated GST model that has been largely discredited. The items for exclusion listed in this Proposition are essentially lifted from the U.K. model.

The Beneficiaries of the Proposed Zero-Ratings

If the U.K. system is so bad, why have the U.K. and, to a lesser degree, other European countries, implemented such a raft of exemptions and zero-rating?

The answer is that the exemptions and zero-rating are an attempt to make the tax more progressive. It is for this same, laudable reason, that P.169/2007 is proposing additional exclusions to GST. But do these exemptions actually make the tax more progressive?

Table  3 below compares the impact of the broad-based tax as currently agreed by the States, with the impact of the GST with the proposed exclusions on food and newspapers, for different levels of household expenditure:

GST Paid as a Percentage of Household Expenditure by Quintile

Table 3

To explain the table, each quintile represents each 20% of households in Jersey, with quintile  1 being the bottom 20% of household expenditure; the poorest, and quintile  5 the top 20%; the richest.

What the table shows is that the simple, broad-based GST as currently proposed will equate to 1.8% of the expenditure of low income households and 1.6% of the expenditure of high income households. All households will pay almost exactly the same proportion of their expenditure in tax.

The proposed exclusions have only a tiny impact on making the tax more progressive.

It is easier to see the variations in cash terms, where the differences are even more marked:

Cash Impact of Zero-Rating on Household Expenditure by Quintiles (£)

Quintile 1 2 3 4 5

Foodstuffs -43 -52 -67 -93 -121 Newspapers, books and magazines -6 -6 -9 -11 -17

TOTAL -49 -58 -76 -104 -138 Table 4

The lowest income households, although in theory £49 worse off, will not in fact save £49 from the zero-rating of food, newspapers and books as the agreed further enhancement to Income Support in respect of GST will already reimburse the full cost of GST to them. The next group, just above Income Support, would save on average £58 a year, and middle income households would save £76 a year, whilst households on the highest incomes would save £138 a year.

The gross loss of tax revenues from these exclusions will be £3.5 million. This loss, however, would be offset by a saving in income support, of £400,000, which would no longer be needed to insulate low income households from GST on food, books and newspapers. Accordingly, the net loss of tax revenues from the proposed exclusions would be £3.1 million. Of this loss of £3.1 million, nothing would go to those on the lowest incomes, just over £1 million would go to those on low-to-middle incomes and £2 million would go to households with high incomes.

Problems of complexity

The proposals to exclude basic foodstuffs, books and newspapers, although superficially simple, cannot be described as "simple exemptions". The exclusions are based on the U.K. VAT model which is regarded internationally as one of the most complex systems in the world and is even non-compliant in terms of the European Union (EU) directives on VAT harmonisation. The following analysis has attempted, where possible, to follow the U.K. VAT interpretation on liability, but the difficulties, even after over 30 years of live tax experience, are striking.

The U.K. exclusion for zero-rating "food" includes 4 general items as sub-categories:

1. F o o d for human consumption
2. A  n imal feedstuffs
3. S e e dsof plants
4. L i v e animals.

There is no specific legal definition for food (but it includes drink) – "it is what the average person would consider it so" but it excludes catering and a list of 7 excepted items (including ice-cream, confectionery) and a further 7 items overriding the exceptions.

Retail shops selling food, confectionery, beverages and other household items will have to identify, for every individual item they sell, whether or not it is subject to GST, and they will need to maintain sophisticated systems to collect and account for the tax. However, in many ways the compliance impact will be easier on retailers of pre-priced/pre-packed food products imported from the U.K.

From the list above it easy to see that many other business sectors will be affected. Hotels, cafés, restaurants, takeaways, and sandwich shops would have different rates of GST for food (hot soup, sandwiches, cereal bars and apples) which will vary yet again depending whether they are consumed on, or off, the premises. Bakeries will have to determine the liability of many products – biscuits and cakes are zero-rated as food but confectionery is taxable. Chocolate-chip biscuits are zero-rated if the chocolate chips are included in the dough or pressed into the surface. Chocolate shortbread biscuits are taxable. Even cake decorations take on different liabilities – chocolate chips are zero-rated whereas chocolate buttons and flakes are taxable.

Pet shops and garden centres will be affected. Animal feedstuffs are zero-rated but pet food is taxable. But rabbit food is zero-rated whereas guinea pig food is taxable. Hay and straw if sold as animal feed would be zero-rated but taxable if sold as bedding. Seeds and plants grown for human consumption or animal feedstuffs are zero- rated – grass seed is zero-rated, but not if pre-germinated and turf is taxable. Flower plants and seeds are taxable, other than specifically listed edible varieties.

Farmers, butchers and fishmongers will also be affected. Meat and dairy animals would be eligible for zero-rating as would rabbits (other than ornamental breeds) even if kept as pets. Honey bees would be eligible for zero-rating but bumble bees are taxable.

Increasing the complexity of the tax, with mixed rates, provides in-built opportunities for error, or, worse, fraud by miscoding whether goods sold are subject to tax or zero-rated. Not only does this reduce revenue yield it also requires the States to employ additional staff to monitor compliance.

Every type of exclusion in terms of supply (goods and/or services) presents a different challenge but international experience shows that any system with mixed liability goods will present difficulties even at the very start of the supply chain. Under the current system designed for Jersey all imported goods are taxable – this has been welcomed as clean and simple. The only problems likely to be encountered are with valuation. However, as soon as any type of goods are excluded there is a great potential for mis-description (both accidentally and deliberately).

Problems would also occur further down the supply chain in identifying taxed and untaxed goods at the point of supply whether by segregation (on tax invoices), or at the point of consumption (using retail schemes).

The added complexity would ensure many more rulings having to be made, requests for extra-statutory concessions, and appeals before independent Commissioners of Appeal – which would all take research, time and care to prepare. There would have to be more control visits by Income Tax auditors to traders' premises to ensure the increasingly complex GST regime is being accounted for correctly. Any discovery of under declaration would lead to an assessment notice, and possibly penalties, which again would have to be subject to appeal. This would all lead to a spiral of control visits/compliance/rulings/appeals which would be time-consuming and contentious. This would in turn also make the tax less acceptable to consumers and businesses, and could lead to further policing costs. Consequently, it could significantly add to the costs of the business, which would be passed on to their customers. This is not theory; this is what happens in the U.K. now.

Excluding the proposed items from GST will almost certainly lead to the delay of traffic of incoming goods at the port. Imported consignments with mixed descriptions would need to be processed to check and assess what goods are correctly applied, or excluded, to GST. This will undoubtedly require additional staffing and lead to a delay in the processing of incoming goods. This could lead to possible stockpiling of goods at the harbour and the airport, a concern of which the Jersey Chamber of Commerce has indicated must be avoided.

Impact on importers and Customs

Under the current GST Law, all goods imported into Jersey are potentially taxable (there will be a de minimis value below which goods will enter freely). Under these circumstances our proposed clearance procedures are simple and have been welcomed by the main importers and Chamber of Commerce. Any potential problems likely to be encountered by Customs are mainly limited to under-valuation.

However, if the proposed exclusions are approved, circumstances will be very different. Goods imported into Jersey will either be subject to 3% or 0% GST. Firstly Customs will be required to maintain an up-to-date and accurate Tariff to include a GST liability indicator for all commodities. Inevitably this will lead to additional problems of mis-description and perhaps a combination of mis-description and under-valuation (deliberately or by genuine error).

This will undoubtedly require additional staffing for Customs, and for non-GST registered importers has the potential to lead to delays in the clearance of incoming goods. This is something we have all been desperately trying to avoid.

It has been asserted that the States has already lost the aim of a simple tax by agreeing to a number of exclusions including education, child care, and medical supplies. This is incorrect. The exclusions to date have primarily been for services, rather than goods, provided from dedicated establishments, such as schools, and the tax system remains simple. On the other hand the administrative implications for processing the proposed exclusions of certain goods on imports, and for businesses having to account for mixed rates of tax, are enormous. The exemptions for goods now being proposed would dramatically increase the cost, for both the States and business, of accounting for and collecting the tax, and ensuring compliance.

Exclusion Creep

Once a start is made to exclude items on the basis that they are an essential spend of the less well-off, or because they are a virtuous activity, there will be no real defence to continued exclusion creep. If newspapers and books are excluded, then why not also petrol, electricity and domestic oil? If excluding children's clothes, then why not equally essential adult clothes? And lightbulbs? And beds and chairs? And toothpaste and soap? Where would it end ?

Other countries

It is true that some countries have a number of exclusions, or reduced rates, in their GST or VAT, but it is generally accepted that the most successful application of these taxes is in countries that have a simple broad- based tax, with a single rate and a high threshold. In fact, the countries generally held up as a good model for a GST are New Zealand and Singapore, where exclusions have been kept to a minimum.

It is also interesting to note that the proposed exclusions are generally subject to GST/VAT at rates above 3% in most of the E.U. member states, as shown in the attached Annex.

Analysis of the Annex reveals that most EU jurisdictions (22 out of 25 jurisdictions listed) apply VAT, to some extent (full rate, standard rate or reduced rate), to foodstuffs. With the exception of Eire and the U.K all the countries listed apply VAT to children's clothing, indeed at the full standard rate. VAT is applied to books and newspapers across all jurisdictions except the U.K. Only medical and dental care are widely treated across the EU as exempt from VAT and this is an item that has now also been excluded from the treatment of GST in Jersey.

Further exclusions will complicate what was intended as being the modern "simple" variation of a GST system. The simple system was the unanimous choice following a period of public consultation conducted by Crown Agents in late 2004. It is not just Jersey that has reached this conclusion. Any country currently going through the same process and considering the implementation of a GST/VAT arrives at the same preferred option. For example, Malaysia, Bahamas and Dubai are at different stages on the GST roadmap but have all discounted the U.K./European style system in favour of the New Zealand /Singapore simple model.

Staffing and Financial Consequences

The proposed exclusions will reduce the gross revenue yield by £3.5 million, and the net yield, after savings in income support, by approximately £3.1 million.

It is difficult to quantify the exact additional administrative costs of these exclusions. However, a reasonable approximation based on U.K. experience is that 8 additional staff will be required and the extra payroll, social security, IT, accommodation and other costs would be approximately £800,000 a year.

Therefore, the total cost to the States of these exclusions, in terms of loss of revenue and increased administration would be of the order of £3.9 million.

The extra cost to business is difficult to quantify, but whatever it is, is likely to be passed on to consumers, rich and poor alike, in increased prices.

Senator Shenton's Incorrect Assertions

In his accompanying report to P.169/2007 Senator Shenton makes a number of incorrect assertions, which really have to be refuted. Firstly, he states that "a tax on food would be highly regressive". As this paper has demonstrated, this statement is incorrect; the tax is slightly regressive.

Senator Shenton also states that financial services in Jersey will be exempted from GST. Again this is simply not true. Whilst it is true to say that generally speaking financial services around the world are exempted from VAT/GST, proposals will shortly be presented to the States which ensure that the finance industry in Jersey contributes some £5-£10 million to the GST tax take.

Senator Shenton also uses the Australian Household Expenditure as background for his arguments. The question should be asked as to why he has not used the Jersey Household Expenditure Survey (JHES) published last year. Surely this data would provide more credible evidence in relation to the possible effects of GST in the Island?

Furthermore, his statement that "Expenditure per person on food scarcely differs with Household income" is patently untrue. If Senator Shenton had taken the time to look at the recent Jersey Household Expenditure Survey rather than the Australian report he would notice that the average weekly household expenditure of the lowest quintile spent on food and non-alcoholic drinks in Jersey is £36.60 per week, whereas that of the highest quintile

is £100.60 per week.[1]

Senator Shenton also suggests that Jersey is no longer a low tax economy. Nothing could be further from the truth. As Chart  1 below demonstrates, Jersey has one of the lowest percentages of taxation as a proportion of Gross Domestic Product (GDP) in the world.

Chart 1

60 50 40 30 20

Tax as a percentage of GDP

10 0

UK

USA Italy

Jersey Mexico Korea Japan Ireland AustraliaTurkey CanadaPoland PortugalGermanySpain Greece Hungary Iceland Austria France NorwayFinlandBelgiumDenmarkSweden EU 19 EU 15

Switzerland NetherlandsLuxembourg OECD total

New Zealand OECD PacificOECD Europe

Slovak Republic Czech Republic

OECD America

OECD nations and Jersey

Senator Shenton's Report also states that Jersey has managed perfectly well for many decades in taxing some goods and not others without the need for massive and expensive bureaucracy. He refers to excise duties collected by  Customs  and  Immigration.  This  again  displays  a  basic  misunderstanding  of  tax  systems  and  their administration. Excise duties are generically considered to be high yield/low cost. The reasons for this are the ease of definition; imposed mainly on goods; it is a single stage tax only (collected at import or domestic production) and involves few taxpayers (in Jersey we are dealing with less than 30 Excise payers of any significance). GST/VAT is a completely different tax system levied on imports and all levels of the supply chain (multi-staged); involves  supplies  of  goods  and  services;  definitions  can  be  difficult;  taxpayers  can  be  regular  payment  or repayment and we are dealing with larger numbers (for Jersey we estimate up to 2,000 GST registered taxpayers).

Conclusion

The Fiscal Strategy should and must be taken as a package. The individual elements of taxation, economic growth and, crucially, Income Support have been designed to complement each other. Whilst GST by itself is regressive, the package overall produces a progressive effect.

Furthermore, the Income Support proposals approved by the States will insulate those on low incomes from the effect of GST. It is universally accepted that this is a far more effective way of protecting the less well-off from the effects of GST than blanket exclusions.

In summary – the States has already decided on 3 separate occasions that it wants a broad-based GST at 3% with few exclusions. The States has repeatedly made this decision in the full knowledge that a broad-based GST would be, on its own, slightly regressive. There were good reasons for this:

GST was never meant to be considered in isolation, but instead always intended to be part of a tax package with the progressive elements being 20% means 20%' and Income Support;

broad-based GST would enable a low 3% rate;

exclusions do not make the package significantly more progressive;

it may not be able to retain the 3% rate indefinitely if exclusions are agreed;

exclusions significantly increase the complexity of the tax and hence the proportion of the tax that is

spent on administration, plus adding to business overheads which, ultimately, the public will end up paying for through increased prices;

granting exclusions from GST encourages yet more calls for further exclusions – exclusion creep. Recommendation

The Minister for Treasury and Resources therefore urges States Members, for the reasons given above, to once again emphatically reject the proposals for exclusions from GST outlined in P.169/2007, and support the Minister's proposals for a further increase in tax exemption thresholds.

ANNEX

EU VAT Rates (also Singapore and New Zealand)

Note:  Where there is more than one rate for any group then the lowest application rate is shown. Sources: European Commission (DOC/1803/2006) and New Zealand and Singapore Government websites.

_______________________________________________________________ Re-issue Note

This comment is re-issued because the attached Appendices 1 and 2 were not included with the text of the comment when originally submitted to the States Greffe for publication.

[1]

Page 22, Report on the Jersey Household Expenditure Survey 2004/05.