Duplicate services are defined in the 2017 Transfer Pricing Guidelines (TPG) of the OECD as “activities undertaken by one group member that merely duplicate a service that another group member is performing for itself, or that is being performed for such other group member by a third party.” 
Those activities are expressively labeled as not being intra-group services and therefore they are not chargeable to the group entity deemed to receive the duplicate services.
The OECD mentions two exceptions where the services will indeed be duplicated but this duplication is justified and thus chargeable: “An exception may be where the duplication of services is only temporary, for example, where an MNE group is reorganizing to centralize its management functions. Another exception would be where the duplication is undertaken to reduce the risk of a wrong business decision (e.g. by getting a second legal opinion on a subject).”
Therefore the two types of exceptional situations relate to a time criteria, the fact that the duplication could be justified if temporary, or to a business criteria, from which it derives that the duplication can be legitimate if driven by risk management considerations.
Once both exceptions have been taken into account, the TPG provides that the analysis must be completed on a factual basis. The examination must detail the nature of the services and the reasons why the entity has or has not received duplicate services. To demonstrate that services have not been duplicated, the company needs to show that “the intra-group services are different, additional, or complementary to the activities performed in-house.”
In parallel, the US definition of duplicative services uses different terminology and is more connected to the concept of benefit: “If an activity performed by a controlled taxpayer duplicates an activity that is performed, or that reasonably may be anticipated to be performed, by another controlled taxpayer on or for its own account, the activity is generally not considered to provide a benefit to the recipient, unless the duplicative activity itself provides an additional benefit to the recipient.” 
It must be noted that the 2017 UN Practical Manual on Transfer Pricing for Developing Countries also defines duplicate services, in a similar fashion as the TPG, but it offers supplementary guidance compared to the OECD, that will be developed further on.
What are the main issues with these activities?
In the public comments on the future revision of Chapter VII of the TPG on intra-group services, commentators globally agree on the fact that duplication is undesirable for MNEs. For instance, Deloitte states “Groups are incentivised to remove inefficiencies in order to maximise returns for shareholders, and any short term duplication would be identified and stopped as part of normal efficiency and management procedures.”
This argument shows that duplicated services are eliminated as soon as they are identified by MNEs and consequently are rare. However, although duplication cases are infrequent, the topic has led to tensions between taxpayers and tax administrations. Many commentators such as Duff and Phelps and EY criticize some tax authorities for automatically considering that where a service recipient receives services with a similar name by different service providers at regional and local levels of the MNE, for example, they are duplicated. They consider that some tax administrations reach this conclusion without fully comprehending the operating model of the MNE and how the service charges are distributed among entities at central, regional and local levels. It is then to the taxpayer to prove the contrary, hence the significant administrative burden.
Moreover, as highlighted by Taxand, “in practice the difficulty is to identify such duplicated activities and to properly distinguish duplicated activities from activities that complement each other and/or constitute additional activities.” Therefore, the main issue with duplicate services is that, although they are unwanted costs for MNEs, they imply a significant administrative burden for the taxpayer and can lead to tensions between tax administrations and taxpayers.
Possible suggestions to Policy Makers
What can be done so the administrative burden related to duplicated services is no longer so considerable for taxpayers? Should the OECD test on duplication of services be reshaped?
To provide a partial answer to the first question, the issue of multi-level services that leads to tensions between MNEs and tax administrations should be addressed. In the public comments, NVB suggested that “Guidance on the role of regional service centers would be useful, as tax authorities often fail to understand that head office may delegate part of its activities to a “regional head office”.
Guidance is always a valuable addition, nonetheless, in this case, a paragraph similar to the one contained in the UN TP Manual might be part of the solution: “At times an MNE group may engage in service functions which have the same name but the functions are performed at different levels and therefore do not involve duplication. These functions may be carried out at group, regional or local level.” This statement incentivizes tax administrations to further analyze the nature of the multi-level services and the operating model of the MNE group. It forces them to be more thorough in their evaluation and not automatically consider the services as duplicated in these specific cases.
Additionally, according to some commentators, such as the Capital Markets Tax Committee of Asia and Duff and Phelps, a way to reduce the administrative burden would be to impose a minimum level and specific type of documentation to be prepared by the taxpayer to demonstrate that activities are not duplicated services. It would lead to less tension between MNEs and tax administrations as the latter would less challenge the MNEs conclusions that would be based on proper documentation. The OECD would evidently need to provide guidance on the required documentation, that could, for instance, contain the following elements: identification of potential duplicative services; contracts and agreements; description of the services rendered or received; reasons why not to be considered duplicative services; expected benefits from the service.
On the other hand, to relieve taxpayers from this administrative burden, other commentators suggested that the burden should be reversed and imposed on the tax administrations. As remarked by Duff and Phelps: “given the rarity with which duplicative services are surely actually provided, once a satisfactory level of evidence has been amassed, the burden should be on the tax administration to demonstrate that services are duplicative, rather than on the taxpayer to demonstrate that they are not.” They also all agree on the fact that, if the burden is reversed, the taxpayer must have previously submitted sufficient information and documents to facilitate the tax administrations’ analysis.
Furthermore, the duplication test stated in the TPG could be reshaped to achieve simplicity and reduce the administrative burden that lies on taxpayers. Instead of having a stand-alone test, clear guidance linking the benefit test with duplicate services is necessary. This position would entail that the benefit test is satisfactory and sufficient to determine whether or not duplication is occurring. Indeed, duplicated activities are not deemed as intra-group services, meaning that they do not provide any additional benefits to the entity. As a matter of fact, this position is close to the US regulations. Indeed, it seems from these regulations that the policy makers stood for a duplication test closely linked to the benefit test: “the activity is generally not considered to provide a benefit to the recipient, unless the duplicative activity itself provides an additional benefit to the recipient.” Thus, if the taxpayer can demonstrate that the activity of the alleged duplicate service provides an additional benefit, it is not a duplication.
Moreover, the UN TP Manual goes in the same direction as it states that “There are some circumstances in which duplication may provide an associated enterprise with a benefit if an independent party would have been willing to pay for the duplicated services in similar circumstances.” In this paragraph, a reference is made to “the willing to pay test” which is also contained in the benefit test of the TPG.
While the OECD uses vague terms such as different, additional or complementary to determine whether activities are duplicated or not, both the US and the UN tend to the conclusion that duplication is not occurring if the services are providing a benefit to the entity.
To conclude, duplicate activities are not intra-group services, unless the duplication is temporary or related to risk management considerations. It follows that they are not chargeable.
The administrative burden to demonstrate that activities are not duplicated relies on the taxpayer and this additional and excessive burden has led to tensions between MNEs and tax administrations. To reduce this burden, documentation could be prepared by the taxpayer but the administrative burden would still lie on the taxpayer. An alternative option would be that the burden could be reversed and laid on the tax authorities instead, if sufficient information has been submitted by the taxpayer. In our opinion, the latter option would indeed relieve the taxpayer from this burden only to pass it on to the tax administrations even though the MNE is better able to demonstrate the absence of duplication due to its more comprehensive understanding of its operating model. Thus, standardized documentation could indeed reduce the administrative burden and appease the relations with the tax administrations when it comes to duplicated activities, but the administrative burden to demonstrate that duplication is not occurring should still lie on the taxpayer in our opinion. The documentation could, for instance, contain the following elements: identification of potential duplicative services; contracts and agreements; description of the services rendered or received; reasons why not to be considered duplicative services; expected benefits from the service.
Nonetheless, no matter if this burden was borne by the taxpayer or the administration, it would still be disproportionate in view of the rarity of duplication. In every case, it is as tortuous and complicated to ascertain whether activities are duplicated or not. In our opinion, the most effective way to ease the taxpayer would therefore be to reshape the duplication test to more closely link it to the benefit test, as it is done in the US regulations and UN TP Manual.
All views expressed are personal. The author would like to thank Mr Stefaan De Baets (PwC) and Prof Dr Vikram Chand for commenting on the draft versions of this article.
 OECD (2017), OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations 2017, OECD Publishing, Chapter VII, Para 7.11.
 OECD Transfer Pricing Guidelines, supra n 1.
 OECD Transfer Pricing Guidelines, supra n 1.
 26 CFR § 1.482-9 – Methods to determine taxable income in connection with a controlled services transaction.
 United Nations, Practical Manual on Transfer Pricing for Developing Countries (United Nations 2017), B.4.2.18.
 OECD (2018), OECD Scoping of the future revisions of Chapter VII (intra-group services) of the Transfer Pricing Guidelines – Comments Received on the Request for Input, OECD Publishing, p. 217.
 OECD Public Comments, supra n 7, p. 79, Deloitte LLP.
 OECD Public Comments, supra n 7, p. 7 and p. 95.
 OECD Public Comments, supra n 7, p. 221.
 OECD Public Comments, supra n 7, p. 83.
 United Nations Practical Manual on TP, supra 6, para B.4.2.19.
 OECD Public Comments, supra n 7, p. 60 and p. 7.
 OECD Public Comments, supra n 7, p. 7.
 US Regulations, supra n 5.
 UN TP Manual, supra n 6, para B.4.2.18.