Skip to main content
Inland Revenue

Tax Policy

Business to business GST neutrality in relation to the cross-border supply of services

Agency Disclosure Statement

This Regulatory Impact Statement has been prepared by Inland Revenue.

The question that this statement asks is what impact should GST have on (a) the supply of services to non-resident businesses; and (b) the charging of tooling costs to non-residents?

Consultation on these issues took place via a government discussion document, GST: Business-to-business neutrality across borders, released in August 2011, which sought feedback on various reform options. Submissions from the private sector were overwhelmingly in favour of the options recommended in this statement.

Public submissions also raised issues related to the detail of the policy that have been taken into account in the final recommendations. For example, submissions confirmed that various options considered in the discussion document (such as requiring a “bond” payment as a condition of registration) would be difficult to comply with in practice and also raised other technical issues that were not considered in the document.

Inland Revenue recognises that the main policy option favoured will impose compliance costs on non-resident businesses if they wish to obtain the benefit of the option. Although other options may impose lower compliance costs, they would impose greater administration costs and risks to the revenue. Further, as the proposed option is not compulsory, businesses do not have to incur the costs if they chose not to register for GST in New Zealand.

There are no other significant gaps, dependencies, constraints, caveats or uncertainties concerning the regulatory analysis undertaken. The proposed option does not impair private property rights, reduce market competition, provide disincentives to innovate and invest, or override common law principles.

Dr Craig Latham
Group Manager, Policy
Inland Revenue

21 May 2012


1. Goods and Services Tax (GST) operates on what is known as the “destination principle”, under which traded goods and services are taxed in the country in which those goods and services are consumed. This principle means that imports are taxed at the standard 15% rate (on the assumption that they will be consumed in New Zealand) and exports are zero-rated (on the assumption that consumption will take place offshore).

2. New Zealand residents that are registered for GST are required to charge output tax on all supplies when the recipient of those supplies is physically in New Zealand at the time the services are performed. This generally provides the right policy result because consumption takes place in New Zealand.

3. GST is designed to be a tax on final consumption. Other than in defined circumstances, it should not impose an economic burden on businesses that form part of the supply chain. This business-to-business (B2B) neutrality is achieved by allowing businesses credit for GST paid, meaning that businesses only return to Inland Revenue any surplus between the GST they charge their customers and GST they incur. If a business incurs more GST than it charges, it is entitled to a refund.

4. This system works well for domestic supplies between GST-registered persons and when goods and services are actually exported and zero-rated. It also works well for tourist consumption, which is generally treated as final consumption in New Zealand and therefore subject to GST. However, it does not work as effectively when a non-resident business comes to New Zealand to receive services.

5. Although a non-resident business is currently entitled to register for GST, its ability to access credit for GST incurred is limited. The result is that a non-resident business that receives services in New Zealand can bear New Zealand GST as a real economic cost, despite the fact that the service received is a genuine business expense that it uses for making supplies to final consumers.

Example: A non-resident airline sends a trainee pilot to New Zealand for part of her training. At the conclusion of her training she is bonded to the airline for five years. The services are received in New Zealand so GST is incurred. Although the training costs are a business expense, because the airline does not operate in New Zealand, it cannot claim back the GST - even though a New Zealand airline in the same position could.

6. When non-residents are not significantly price sensitive in relation to the particular services, or when New Zealand businesses offering the services have a distinct competitive advantage, this taxing of non-resident businesses might not be a significant concern. However, in addition to evidence presented by some affected taxpayers, various factors indicate that the imposition of GST could be a contributing factor in New Zealand businesses being placed at a competitive disadvantage:

  • Non-resident businesses are likely to be increasingly price-sensitive in the current economic environment;
  • Comparable jurisdictions, such as Australia, Canada and the European Union, already have systems in place to prevent their GST (or equivalent) being an economic burden on non-resident businesses; and
  • Although there is already a mechanism whereby non-residents can structure around this problem (that is, by incorporating a New Zealand subsidiary and registering that company for GST), there are relatively heavy compliance costs associated with this.

7. A separate, but related, issue exists in respect of tools and similar equipment that are used in New Zealand exclusively to create goods that are exported. New Zealand-resident manufacturers are required to charge GST on these tools despite the fact that they are only used to create goods that are themselves zero-rated for GST purposes. Again, this arguably places New Zealand manufactures at a competitive disadvantage to manufacturers operating in jurisdictions that have specific rules to remove any GST impost in this situation (such as Australia and the United Kingdom).

Example: A non-resident contracts a New Zealand manufacturer to make a branded vending machine for export. The non-resident is charged for the finished machines and the special tools required to create the unique product. The New Zealand manufacturer must charge GST on the tooling costs despite the fact that they are only going to be used to make the exported (and zero-rated) vending machines. The non-resident cannot claim back the GST.

8. Maintaining the status quo in these areas would protect the existing revenue generated through supplies to non-resident businesses. However, this option also has the potential to disadvantage New Zealand businesses and manufacturers with significant non-resident business client bases. Affected businesses have noted that the discussion document proposals would reduce competitive disadvantages and increase the international competitiveness of New Zealand businesses.

9. The status quo is also out of step with OECD best practice, which says that non-resident businesses should have access to the same GST result offered to a comparable resident business that acquires the same goods and services. As stated above, other OECD countries have rules that reduce or eliminate their tax impost. Other countries also operate on the basis of reciprocity – meaning that New Zealand businesses are currently denied value added tax (VAT) refunds in some European countries because New Zealand would not offer that country’s businesses a refund of GST incurred here. In this way, the status quo may discriminate against non-resident businesses and penalise New Zealand businesses that incur VAT in some European countries.


10. The question to be answered is therefore what impact should GST have on (a) the supply of services to non-resident businesses; and (b) the charging of tooling costs to non-residents?


11. The objective of this reform is to provide greater GST neutrality for non-resident businesses that receive services in New Zealand and for non-residents that incur tooling costs in New Zealand. The desire for neutrality would need to be balanced against compliance costs imposed on business (both resident and non-resident), administration costs for Inland Revenue, and fiscal costs in terms of GST foregone.


It is recognised that the most desirable outcome for the supply of services to non-resident businesses (referred to be below as “exported services”) might not necessarily be the best for tooling costs. This statement therefore considers each issue in turn.

(a) Exported Services

12. A Government discussion document, GST: Business-to-business neutrality across borders, was released in August 2011 seeking public submissions on these issues. Three viable options were identified and considered as means of enhancing business-to-business neutrality for supplies of services to non-resident businesses:

  • zero-rating certain supplies;
  • relaxing the restrictions on non-resident businesses for claiming input tax; and
  • a refund system for non-resident businesses.

13. The first option builds on rules already in place under the Goods and Services Tax Act 1985 (GST Act). The second and third options involve enhancing neutrality though adopting problem-specific systems that are already used overseas. The advantages and disadvantages of each system set out below largely reflect those identified the August 2011 discussion document.

Option 1: Zero-rating (not recommended)

14. The GST Act already has rules related to when exported services can be zero-rated. One option is therefore to extend these rules to cater for circumstances when a non-resident business receives the relevant services in New Zealand.

15. An industry-specific approach would be inconsistent with our GST system’s broad base and could create distortions across industries. A zero-rating solution would therefore need to be broad in its approach so as not to diminish economic efficiency.

Advantages of zero-rating

16. Zero-rating is consistent with the destination principle and with how the GST Act currently deals with goods and services that are exported. Having a system that aligns the treatment of all exports is simple and transparent.

17. Zero-rating is a mechanism that is familiar to New Zealand businesses, and would impose minimal compliance costs on non-resident businesses. Zero-rating would also result in minimal administrative costs for Inland Revenue, as the reporting requirements would fall on existing registered persons.

Disadvantages of zero-rating

18. Capturing the type of services to be zero-rated would be difficult and would likely create areas of contention around what was, or should be, zero-rated.

19. Irrespective of how the relevant rules were drafted, there may be “gaps”, which would create distortions through GST being a cost to a non-resident business when the same GST charged to a resident business may be able to be deducted as input tax.

20. An expanded zero-rating system may also push potentially large compliance costs onto New Zealand suppliers. The New Zealand-resident supplier would be required to identify each customer and determine the appropriate GST treatment on a transaction-by-transaction basis. For suppliers that do not deal exclusively with non-resident businesses, this could be extremely difficult to administer. A New Zealand service provider would have to establish not only that their client was a non-resident, but also that they were consuming the services purely for business purposes.

21. In addition, when a supply is zero-rated, the output tax that would have been chargeable on that supply is immediately “lost” from the GST system. This would make any mistakenly zero-rated supply difficult for Inland Revenue to identify and recover GST from. The GST Act would make the supplier in such cases liable for output tax not charged on the supply, which would be another cost to New Zealand businesses.

22. If a broad solution involved listing activities to be zero-rated, any rules would also be vulnerable to shifts in behaviour in the types of supplies made. A definition of certain activities would be fixed in time and may not be adequately future-proofed. Such rules would be more likely to require future amendments than general, process-based legislation.

Option 2: Enhanced registration system (recommended)

23. A system termed “enhanced registration” would be a variation on the current system and similar to the rules that operate in Australia. Under this system, a non-resident business would retain their existing right to register for GST, but this would be coupled with more generous rules around claiming input tax deductions.

Advantages of enhanced registration

24. Non-residents would be operating under a similar system to residents, thereby enhancing the neutrality objective. If a non-resident business made no taxable supplies in New Zealand, they would be entitled to a refund of all input tax related to their offshore business activities (subject to restrictions to ensure that non-resident businesses are not able to claim more input tax than a comparable New Zealand-resident business). If the non-resident were to make supplies in New Zealand, as a registered person, they would be required to account for output tax on those supplies.

25. The system would be optional for non-residents. A resident could choose whether or not to register, so compliance costs associated with the system could be avoided by not registering.

26. An enhanced registration system would be complementary to existing Inland Revenue systems, as a non-resident business would be required to complete and file GST returns in a similar way to a New Zealand-resident registered person.

27. An enhanced registration system would be similar to the system used in Australia. Having broadly similar rules would make it easier for Australian businesses to operate in New Zealand and for third-country businesses to operate in both jurisdictions.

28. The GST system already contains most of the registration and input tax deduction provisions necessary to make an enhanced registration system work. Even though introduction of such a system would result in more GST-registered persons, it would not require complex new legislation and systems to be introduced.

Disadvantages of enhanced registration

29. In order to access input tax deductions, non-resident businesses would need to incur the compliance costs of registering for GST and filing periodic returns. These costs would not necessarily be large, particularly given developments in online registration and return filing. However, some non-resident businesses may consider it necessary to engage a New Zealand advisor to help with ongoing filing requirements.

30. Asking non-residents to fully engage in the New Zealand GST system could be seen as a deterrent rather than just an administrative step towards eligibility for a refund. For example, businesses could be concerned that GST registration could give rise to New Zealand income tax issues.

31. Although Inland Revenue has systems for allowing non-residents to register, administration costs would nevertheless increase in line with the increased number of registered persons.

32. New Zealand’s current legislation has the basic mechanisms in place for non-resident businesses to register for GST, claim input tax and receive refunds. However, some additional legislative and systems safeguards to protect the tax base would be necessary and this could increase the complexity of this option.

Design features

33. The discussion document also recommended that any enhanced registration system should incorporate features designed to ease administration and protect the tax base, including:

  • Requiring the non-resident to satisfy the Commissioner of Inland Revenue (“Commissioner”) that they are either registered for a comparable tax in their own jurisdiction or would be eligible to register if they were a New Zealand business, including that their supplies are over the registration threshold of $60,000;
  • There being a minimum refund threshold of $500 in the first period of registration to reduce administration costs;
  • The Commissioner having three months to process refunds for non-residents (compared with the 15 working days that the Commissioner has for New Zealand residents);
  • Requiring non-residents to account for GST on a payments accounting basis;
  • Non-resident businesses that on-supply New Zealand services to non-registered persons not being eligible to register.

Option 3: Direct refund system (not recommended)

34. A direct refund system would be largely similar to the way refunds are dealt with under the VAT system of the European Union (EU). Under this system a non-resident business is entitled to a VAT deduction if the person does not make taxable supplies within the relevant EU member state. The non-resident business completes a separate application form seeking the deduction in the form of a direct refund for VAT incurred. The refund claim is usually made after the end of a defined period.

Advantages of a refund system

35. Like an enhanced registration system, a refund system has the potential to create full neutrality between resident and non-resident businesses.

36. A refund system is relatively easy to understand and is likely to be familiar to multi-national businesses, particularly those that incur GST or VAT costs in Europe.

37. A refund system can also be accessed on the basis of the need of the individual business, thereby eliminating the requirement for upfront registration and ongoing compliance costs associated with regular return filing.

Disadvantages of a refund system

38. A refund system could be more likely to give rise to fraudulent refund claims than a system that requires non-resident businesses to register as a pre-requisite to obtaining a refund.

39. The administration costs of this option are likely to be prohibitively high for Inland Revenue. New forms would be needed, and staff hired or retrained to ensure applications were processed and monitored appropriately and refunds issued in a timely manner. More significantly, a refund system may require new IT systems to be developed.

40. Such developments could be seen as being counter to Inland Revenue’s programme of streamlining customer interactions and IT systems.

(b) Tooling costs

41. In the manufacturing sector, there is a specific pricing model that is used to quote for, and undertake, work of a tailor-made nature. This system involves quoting for the production of a certain quantity of goods (either in bulk or on a per unit basis) and quoting separately for the “tooling costs” associated with completing the order to the specific requirements of the customer. As the name suggests, these tooling costs cover expenses associated with creating or adapting tools that can only be used in fulfilling the particular order.

42. This separation of costs is driven by the customer’s desire to be the “owner” of the specific tools, even though they have no intention of ever taking delivery of them. This is important for the customer because it prevents the manufacturer from using the tools in a future product, thereby reducing the possibility of copy products appearing on the market.

43. However, this pricing structure creates a problem from a GST perspective when the manufacturer is resident in New Zealand and the customer is non-resident. The actual goods being produced will be exported and therefore generally able to be zero-rated. By contrast, the tools will be used exclusively in New Zealand and will not be physically exported. Because the tools are not exported, none of the existing zero-rating rules apply and the tooling costs will be subject to GST at the standard rate.

44. This is not a problem unique to New Zealand. Australia and the United Kingdom are among the countries that have specific rules that zero-rate the supply of tools used exclusively to produce exported goods.

45. Two options were considered to address this issue:

  • Allow the situation to be addressed by an enhanced registration system; or
  • Introduce a specific zero-rating rule for tooling costs.

Use an enhanced registration system (not recommended)

46. The introduction of an enhanced registration system could provide a neutral outcome. This would be achieved by the non-resident business client of the New Zealand supplier registering for GST and claiming back as input tax the GST on the tooling costs. Although the enhanced registration system is seen as the best option to address issues around the cross-border supply of services, as noted above, it will impose some compliance costs. If a simple, targeted and effective zero-rating rule can be implemented to address the issue, this should be preferred because it will reduce compliance costs both for New Zealand and non-resident businesses.

Zero-rating rule (recommended)

47. The tooling costs issue is a specific problem related to goods. Existing legislation in Australian and the United Kingdom shows that the issue can be effectively addressed through legislative drafting.

48. There is also a current rule in the GST Act that addresses an equivalent scenario. That rule provides that when services are provided exclusively in respect of goods that are exported, those services are able to be zero-rated. The zero-rating solution proposed would create a corresponding rule that would allow goods that are used exclusively to produce exported goods to be similarly zero-rated.

49. Any concerns around revenue loss can be addressed in these circumstances by supporting rules that would impose a GST liability on the New Zealand manufacturer if the tools were used for domestic supplies.


50. The government discussion document GST: Business to business neutrality across borders was released by the Minister of Revenue, Hon Peter Dunne, for consultation in August 2011. Twelve public submissions were received from businesses and advisors.

Exported services

51. Of the twelve submissions, nine broadly supported the introduction of an enhanced registration system, which was the option favoured in the discussion document. One submission favoured a direct refund system and two favoured a broadening of the zero-rating rules.

52. The one submission that supported a direct refund model was of the view that this system would be broadly similar to the enhanced registration system in terms of Inland Revenue’s administration costs. Inland Revenue does not agree with this view and believes that the systems and human resource costs associated with a refund system are prohibitively high.

53. The two submissions that favoured zero-rating were both received from representatives of the aviation training industry, which provided the impetus for considering the issue. Their submissions are consistent with their long-held view about the preferred approach and are based on:

  • Reducing compliance costs for their customers, who would not have to do anything to avoid GST if the supply was zero-rated. As referred to above, we agree that this is an advantage of zero-rating. However, we do not agree that the system should favour non-resident businesses to the detriment of New Zealand registered persons.
  • The supplier would choose to zero-rate depending on the value of the transaction and their relative international competitiveness. We disagree with this point because the imposition or non-imposition of tax is, and should remain, mandatory.

54. In subsequent discussions with one of the submitters, at both an officials’ and Ministerial level, there appears to be acceptance that the policy could also be achieved through the introduction of an enhanced registration system.

55. Submitters that did favour the enhanced registration system also provided feedback on various policy details proposed in the discussion document. Suggestions on these details have been taken into account in finalising policy, particularly in the proposed design features listed above.

Tooling costs

56. On the issue of tooling costs, nine of the twelve submissions directly addressed this point. Of those nine, eight favoured the introduction of a specific zero-rating rule, whereas one preferred the option of the enhanced registration system addressing the issue.


57. The consultation process confirmed that the approach favoured in the discussion document was also the preferred option for affected parties. Submissions also confirmed that some elements of detail of any enhanced registration system would be difficult to comply with in practice (such as requiring a “bond” payment as a condition of registration). Finally, submitters raised other technical issues that had not been considered in the document (such as the GST treatment of migrating companies), which would be incorporated in any legislation.


58. For the exported services issue, it is recommended that an enhanced registration system be introduced, and that this system incorporate the design features listed above.

59. For the tooling costs issue, for the reasons given above, it is recommended that a specific zero-rating rule be introduced, along with a rule that imposes GST in the event that the tools are used to produce products for the domestic market.

Estimated fiscal costs associated with recommended changes

60. Introducing the enhanced registration system is estimated to have a fiscal cost of approximately $10 million per annum. This estimate takes into account the fact that non-residents with significant business interests in New Zealand will have already established New Zealand subsidiaries to ensure their eligibility for refunds of GST incurred.

61. Because data on individual transactions is not kept, it is not possible to estimate the tax revenue costs associated with zero-rating tooling costs. However, as it only affects a small proportion of overall supplies made by the export manufacturing sector, it is not expected to be significant.


62. An enhanced registration system should result in an increased number of GST registered persons and is likely to result in the creation of a new GST registration code specifically for non-residents (with associated amendments to the registration form itself). These factors, as well as the smaller changes necessary to ensure that specific rules designed for non-residents are correctly applied, will increase the costs of administering the GST system and will involve minor systems changes.

63. The indicative administration costs over the five-year forecast period are $1.320m. There are not expected to be any administration costs associated with the introduction of a zero-rating rule for tooling costs.

64. Although the systems changes associated with these proposals are not expected to be substantial, in order to allow time for implementation, an effective date for the changes of 1 April 2014 is recommended. The necessary amendments to the GST Act and the Tax Administration Act would be included in the first omnibus tax bill of 2012. This timeframe would allow for that bill to be enacted and for Inland Revenue to deliver the necessary changes prior to the new rules becoming effective in 2014.

65. The only transitional requirement would be the obligation for those non-residents that are already registered for GST to satisfy the Commissioner of their continuing eligibility to register. It is recommended that affected non-residents have a one-year window in which to satisfy this requirement.


66. Monitoring the effect of these changes will fall under Inland Revenue’s responsibilities under the generic tax policy process (GTTP). The GTTP is a multi-stage process that has been used to design tax policy in New Zealand since 1995. The final stage of this process is the implementation and review stage, which involves Inland Revenue conducting a post-implementation review and identifying any remedial issues. Opportunities for external consultation are built into this stage.

67. In addition, as a new registration code will be created specifically for non-resident businesses, Inland Revenue will have the capacity to monitor refunds provided through the introduction of an enhanced registration system.