Cross-Border Remote Work and the Permanent Establishment
By: Martin van Harten (1) and Frank Mélotte (2)
The Revised 2025 OECD Commentary and Its Implications for Practice
1. Cross-border remote work as a lasting phenomenon
Cross-border remote work is no longer a temporary effect of the COVID-19 pandemic, but a structural component of international labor mobility. Employees expect flexibility, and employers use remote work to attract and retain talent. (3)
As a result, this has long ceased to be merely an HR issue: as soon as an employee lives and works across borders, this has implications for taxation and the applicable social security laws. This raises various questions for both the employer and the employee: Which country is authorized to tax which portion of the salary? Who is responsible for compliance obligations? And, last but not least, does the employer establish a permanent establishment in the employee’s country of residence?
This article focuses on that last question, partly in light of the OECD commentary revised in November 2025. We compare the commentary with practices in the Netherlands, Belgium, and Germany and extend the discussion to payroll taxes, social security, and workations. The focus is practical: when does working from home remain merely a matter of filing a different income tax return, and when must the employer reclassify the employment status for tax purposes?
2. Why the permanent home office setup is relevant
Under certain circumstances, an employee’s residence may constitute a permanent establishment (hereinafter “PE”) of the employer. This is primarily a risk for the employer. If a PE arises, the employee’s country of residence may tax a portion of the business profits, and an obligation to register or withhold payroll taxes may arise under national law. The key point is that the country of residence is then no longer merely the employee’s country of residence, but also the source country for the employer’s business activities.
The assessment is based on the traditional cumulative criteria. There must be a fixed place of business that is of a certain permanence and through which the company’s activities are carried out in whole or in part. (4) Since 2020, Dutch legislation has closely aligned with the treaty-law concept. As a result, a treaty-law permanent establishment (PE) can also become directly relevant for Dutch payroll tax and corporate income tax. This makes the “work-from-home” PE not just an abstract treaty concept, but an operational compliance issue. The term “micro-permanent establishment ” (5) is apt in this context: the material profit interest may be small, while the organizational implications are significant. Consider local registration, tax returns, transfer pricing documentation, payroll setup, and engaging local advisors.
When it comes to working from home, the discussion usually does not center on whether the employee is actually performing work. Nor is it typically disputed that a residence, home office, or other relevant location may, under certain circumstances, be considered a place of business and that regular use may satisfy the requirement of permanence. (6) The real uncertainty lies in whether that location can be attributed to the business. In other words: Is the employee’s private space also a place of business for the employer?
3. The Traditional Framework and the Long-Standing Debate
Under the old OECD commentary, great emphasis was placed on the “availability” criterion: Is the home workplace actually available to the enterprise? That criterion does not appear verbatim in Article 5 of the OECD Model Tax Convention, but was developed in the commentary. Consequently, it took on an almost constitutive function, particularly in work-from-home situations. The question then became whether the employer required the employee to work from home or failed to provide a suitable workplace in the employer’s country. (7)
That approach had a practical advantage: when working from home stemmed primarily from the employee’s personal preference, it was often possible to argue that the residence was not at the employer’s disposal. The disadvantage was equally clear. The assessment shifted toward the wording of company policy and the employment contract, whereas the economic reality may in fact be that the business is structurally conducted through that home workspace. As a result, the old criterion sometimes focused more on the place where work was supposed to be done than on the place where work was actually and structurally performed.
4. The Revised OECD Commentary 2025
In November 2025, the OECD published a new framework for cross-border work performed from home or another relevant private location, such as a vacation home or a family member’s residence. The former paragraphs 18 and 19 of the Commentary on the OECD Model Tax Convention were repealed and replaced by paragraphs 44.1 through 44.21. (8) In doing so , the OECD opted for an update that better aligns with hybrid work and international remote work arrangements.
It is important to note that the new commentary does not replace the classic VI criteria. It provides further clarification on when a residence or other relevant location may be considered the company’s place of business. In this regard, greater emphasis is placed on the employee’s actual behavior and on the employer’s business interest in the employee’s physical presence in the country of residence. The old question of whether the employer requires employees to work from home does not disappear entirely, but it becomes less dominant.
The new system consists of two steps. First, there is a quantitative approach: if the employee works less than 50% of his total working time for the company from home or another relevant location, that location does not, in principle, constitute a vi. (9) At 50% or more, that “certainty” disappears. This is followed by an assessment of all facts and circumstances, in which the commercial reason for being present in the country of residence plays a prominent role. This approach is pragmatic, but it also leaves some questions unanswered. For example, the commentary does not provide an exhaustive explanation of how to handle multiple relevant locations, varying work schedules, or team-based roles. The employer must therefore translate these gaps into policy and documentation procedures on its own.
5. The 50% Threshold
The 50% threshold is attractive because, while it is not a true “safe harbor,” it generally provides sufficient guidance. The assessment covers a twelve-month period that begins or ends within the relevant fiscal year. What counts is the employee’s actual behavior. Contractual agreements and work-from-home policies are only relevant to the extent that they correspond with actual practice.10 A policy that allows a maximum of two days of work from home therefore offers no protection if the employee actually works four days a week from abroad and the employer allows this to continue.
The practical implications are significant. Employers who wish to monitor compliance with the threshold must record workdays and work locations. In many cases, an annual declaration from the employee will not be sufficient. In cases of systematic cross-border work-from-home arrangements, periodic recording of location and working hours is the obvious approach. This is administratively burdensome, especially when the potential tax benefit of a work-from-home arrangement is limited. The risk associated with a work-from-home arrangement therefore often lies less in the material tax costs and more in the obligation to record, report, and provide evidence.
6. What happens when the percentage reaches 50% or more?
Exceeding the 50% threshold does not automatically result in liability. The threshold primarily functions as a one-sided “safe harbor”: below the threshold, the employer is, in principle, safe; above it, the employer is no longer safe. The OECD did not opt for a two-sided safe harbor, under which a vi would always arise above the threshold. (11) This is appropriate, because the nature of the work and the business’s interests can vary greatly.
If 50% or more of the work is done from home, the employer must therefore reassess the facts. What tasks does the employee perform? For whom are those tasks performed? Is the employee’s physical presence in their country of residence important for the market, customers, suppliers, staff, or business assets? Is this a temporary situation or a structural work pattern? And how does the actual work pattern relate to the formal work-from-home policy? It is precisely these questions that make it clear that the new assessment must be structured not only from a tax perspective but also from an organizational one.
The burden of proof also warrants attention. During an audit, foreign tax authorities will examine not only the work-from-home policy but also calendars, travel patterns, expense reports, client files, and CRM systems. An employee who consistently serves clients from Belgium or Germany typically leaves behind more evidence than just a work-from-home log. The employer would therefore be wise to build a strong case: why is work being done in the country of residence, which tasks remain at the headquarters, and why is the local presence—or lack thereof—commercially relevant?
7. The Commercial Rationale Test
The “commercial reason” test is the key qualitative factor. A commercial reason exists when the employee’s physical presence in the country of residence facilitates the employer’s business operations. This may be the case, for example, when the employee visits clients there, develops local sales, manages suppliers, or has access to individuals or resources relevant to the company.(12) The question, therefore, is not whether working from home is pleasant or efficient, but whether the country of residence is, in and of itself, significant to the business.
In contrast, there are situations in which a person’s presence in the country of residence is primarily motivated by personal reasons. Consider, for example, an employee who works from abroad due to a partner, family, or personal preference, even though the position is entirely internal and does not serve the local market. According to the commentary, attracting or retaining staff and saving on office costs do not, in and of themselves, constitute a commercial reason. Occasional customer visits are also insufficient. In the absence of a commercial reason, a VI is not obvious, barring other facts and circumstances.
8. Functions by Risk Profile
The commercial-reason test requires job classification. A back-office position, internal finance role, or IT role without local external contacts typically has a lower risk profile. A sales manager with local clients, a business developer building a market, or a procurement specialist managing local suppliers ranks much higher on the risk scale. Management positions also warrant attention, especially when important decisions are prepared or made from the country of residence.
This shifts the policy on working from home from HR to tax governance. A generic policy—such as a maximum of two days of working from home—is not sufficient. The employer must know where the employee works, how much work is performed there, and what business function is carried out from that location. (13) A sound approach combines a light initial screening with escalation where necessary. As soon as the work pattern becomes structural, approaches the 50% threshold, or involves a commercial function, an individual assessment follows.
9. The Netherlands and Belgium: Useful but Limited Certainty
The Netherlands and Belgium had already concluded a bilateral agreement on the work-from-home provision by the end of 2023. That agreement is based on Article 28, paragraph 3, of the Netherlands-Belgium tax treaty and is intended to resolve ambiguities in the interpretation of Article 5.14 Among other things, it distinguishes between occasional work from home, regular work from home while a workspace remains available at the employer’s premises, and mandatory regular work from home.
The practical implications are significant. If an employee works from home in their country of residence 50% or less of the time over a 12-month period, a vi is not deemed to exist. Nor is the provision of IT equipment, a desk, a chair, or reimbursement for work-from-home expenses a deciding factor. As such, the agreement offers more concrete certainty than the general OECD commentary. The agreement primarily helps employers who have a manageable population of cross-border workers and wish to prevent a single Belgian or Dutch employee working from home from triggering corporate income tax registration. At the same time, this certainty is limited. The agreement does not affect the exceptions to the “vi” concept and the permanent representative rule, and it says nothing about the allocation of the employee’s wages. For the employee, wages for work physically performed in the country of residence remain, in principle, taxable there. (15)
10. Germany, the Benelux Countries, and the Policy Context
Germany does not have a comparable mutual agreement regarding the home office rule. However, the BMF follows a relatively broad unilateral policy: as a general rule, an ordinary employee’s home office does not constitute a Betriebsstätte, even if the employer bears the costs or does not offer an alternative workplace. Furthermore, at the end of 2025, the BMF tightened its stance regarding subsequent OECD commentary. From a German perspective, the 2025 commentary will carry particular weight only to the extent that it can serve as a clarification of existing treaty interpretations. (16)
Alongside the VI discussion, there is the broader agenda regarding cross-border workers. The Netherlands wants to provide tax incentives for working from home in the border region, but structural solutions require treaty amendments or bilateral agreements. A threshold arrangement has been agreed upon with Germany, allowing cross-border workers to work from home for up to 34 days per year without a shift in the right to tax their earned income. A comparable work-from-home arrangement under Article 15 has not yet been reached with Belgium. The issue has been put back on the agenda at the Benelux level, but positions differ. (17)
11. Social Security
The 50% threshold mentioned in the OECD commentary should not be confused with the 50% threshold in the Framework Agreement on Structural Cross-Border Telework. Under that framework agreement, upon request—through an Article 16 agreement and confirmation via an A1 certificate—it is possible to ensure that the employee remains covered by social security in the employer’s country, provided that telework in the employee’s country of residence accounts for less than 50% of total working hours and there is no structural work in a third country. Outside the scope of this agreement, the usual 25% threshold for substantial work in the country of residence applies in multi-country work situations. (18)
Social security therefore requires a separate analysis. The concepts are not the same. The tax VI is an attribution tool for business activities, whereas the social security coverage requirement is tied to the individual. The concept of residence and the relevant time units also differ. An employer may therefore remain outside the scope of tax VI risk, while action is still required under social security law. The reverse is also true: an A1 certificate does not prevent tax-related VI. In practice, this means that tax and payroll departments must conduct their assessments in parallel, rather than one after the other. Apply for an A1 certificate in a timely manner, monitor changes in work patterns, and do not wait until payroll notices the shift. (19)
12. Consequences for the Employee
Although the issue primarily concerns employers, cross-border remote work also affects employees. Under Article 15, wages attributable to work physically performed in the country of residence (more accurately: outside the qualifying country of employment) are, in principle, taxed in the country of residence.20 For the employee, therefore, it is primarily the number of days worked from home that is relevant, not whether the residence qualifies as a vi.
In addition, working from home can affect qualifying foreign tax liability. Foreign taxpayers who wish to retain Dutch personal deductions or home mortgage interest must, in principle, meet the 90% requirement. As more workdays shift to the country of residence, the proportion of income earned in the Netherlands may decrease. This can lead to the loss of deductions and an unexpected tax liability. These consequences for employees do not belong in the VI analysis itself, but must be monitored within the same mobility file. (21)
13. Workation
The new commentary refers not only to the home but also to another relevant location. This brings the concept of a “workation” into the picture: temporarily working from a vacation home, a family residence, or another private location. For the vi, the risk is generally limited in the case of short-term arrangements because they lack sustainability. The commentary cites as an example that working from a rented home for three consecutive months is not sufficiently long-term for a self-employed individual. (22)
That does not mean that workations are tax- or legally neutral. Workdays spent in the workation country count toward wage allocation. For longer stays, issues such as tax residency, social security, labor law, immigration, and local registration requirements may also come into play. A sound workation policy therefore includes a maximum number of days, a list of approved countries, a prior approval process, and a ban on client visits or commercial activities in high-risk countries without a tax review.
14. Consequences Beyond Payroll Taxes
A VI has implications beyond payroll taxes. In the country of residence, an obligation to file an income tax return may arise, with profits attributed to the VI in accordance with the arm’s-length principle and the associated transfer pricing issues.23 In the case of a home-based VI, the attributable profit will often be limited, but the compliance burden may be disproportionate: registration, filing, documentation, local advisors, and correspondence with foreign tax authorities.
15. Does the new commentary apply to old treaties?
A fundamental question is whether the 2025 OECD commentary constitutes a clarification or a substantive amendment. Pursuant to the Supreme Court’s ruling of October 14, 2022, post-treaty commentary can serve only as a supplementary means of interpretation within the meaning of Article 32 of the Vienna Convention, and only to the extent that it constitutes a clarification or elaboration. (24)
In our view, caution is warranted. The 50% threshold and the commercial reasons test were not included in the previous commentary. Furthermore, the now-defunct paragraphs 18 and 19 relied heavily on the criterion of disposition. This suggests that the new approach should not automatically be used as a primary interpretive tool for older treaties. For each treaty, it must be assessed whether dynamic interpretation is required, whether there is a mutual agreement, and how the states involved position the new commentary. (25)
16. Practical Management Measures
Four control measures apply to the practice.
‐ First: determine the work pattern. In accordance with the OECD commentary, the general rule is that if less than 50% of work is performed from home or another relevant location, a permanent establishment (PE) is not, in principle, deemed to exist. For the Netherlands-Belgium agreement, a practical guideline is that if 50% or less of work is performed from home, a permanent establishment is not presumed to exist.
‐ Second: Classify positions based on their risk profile. Commercial positions, local customer-facing roles, and supplier management require a more rigorous review than internal positions.
‐ Third: Establish an escalation procedure. An employee who consistently works from home beyond the permitted limits, is approaching the 50% threshold, or engages in commercial activities in their country of residence must consult with the tax and payroll departments in advance.
‐ Fourth: monitor the interplay of these factors. Check salary allocation, A1 status, qualifying foreign tax liability, workation rules, and immigration law. Written agreements are useful, but not decisive. The actual course of events takes precedence over policy. (26)
17. Conclusion
The revised OECD commentary brings the “work-from-home” test closer to actual reality. The assessment shifts from whether the employer has formally made the home workspace available to the employee’s work pattern and the employer’s commercial interest in the country of residence. That is a step forward, but it does not provide complete certainty. The 50% threshold is, in practice, a useful safe harbor, not an absolute limit. The commercial-reason test remains open-ended, and national approaches continue to vary.
The practical challenge, therefore, is governance. Employers would do well not to block cross-border remote work, but to manage it. This requires tracking workdays, classifying job roles, obtaining prior approval for long-term arrangements, coordinating with social security and payroll departments, and establishing a clear “workation” procedure. Tax considerations must be taken into account early in the decision-making process. Not only when an employee has already been working from abroad for months, but at the time of the request, the contractual agreement, and the periodic review. Those who set this up properly can offer flexibility without unnecessary tax risks. Those who fail to do so underestimate the “kitchen table” as a source of taxable income. (27)
Notes:
M.L.M. van Harten is a partner at Nassau Tax & Global Mobility.
F.E. Mélotte is a partner at Nassau Tax & Global Mobility.
See, among others, M.J.C. Vedder & E. Babuni, “The Challenges of Working from Home: The Kitchen Table as a Taxable Object,” WFR 2023/90, para. 1, and L.H.A. Jacobs & R. Kühl, “Classification of a Home Office as a Permanent Establishment,” MBB 2025/11, para. 1.
Art. 3, paras. 4 through 12, Corporate Income Tax Act of 1969, as applied to Art. 1.10 of the Income Tax Act of 2001, Art. 6, para. 3, subpart a, of the Personal Income Tax Act of 1964, and Art. 2 of the Double Taxation Prevention Decree of 2001. See P.G.H. Albert, “Finally, a Chameleon-Like Concept of Permanent Establishment,” WFR 2019/209.
See, for example, Mădălina Cotruț, “Are Micro Home Office PEs Tackled under the 2025 Update to the Commentary on the OECD Model?”, *European Taxation*, 2026 (Volume 66), No. 2/3.
Supreme Court, October 13, 1954, ECLI:NL:HR:1954:AY4080; Supreme Court, June 15, 1955, ECLI:NL:HR:1955:AY4120; Supreme Court, March 13, 1957, ECLI:NL:HR:1957:AY1589, BNB 1957/144. See also Vedder & Babuni, WFR 2023/90, para. 2.
Commentary on Article 5 of the 2017 OECD Model Tax Convention, paras. 18 and 19 (old). For criticism of the control criterion, see MBB 2026/25, para. 2.
OECD, The 2025 Update to the OECD Model Tax Convention, November 19, 2025; commentary on Art. 5, paras. 44.1 through 44.21. See R.K. Bane, “Update to the OECD Model Tax Convention and OECD Commentary 2025,” NLF-W 2026/1, and S.H.J.P. Adriaans, NLF 2025/2607.
Commentary on Article 5 of the OECD Model Tax Convention (2025), paras. 44.1 through 44.8. See also MBB 2026/25, para. 3.1.
Commentary on Art. 5 of the OECD Model Tax Convention (2025), para. 44.8; see A.I. Harteveld & F.P.G. Pötgens, “Interpretation of the Permanent Establishment Definition in an Era of Cross-Border Remote Work,” WFR 2025/175.
Commentary on Article 5 of the OECD Model Tax Convention (2025), para. 44.10. For more on the choice of a unilateral safe harbor, see MBB 2026/25, paras. 3.1 and 6.2.
Commentary on Article 5 of the OECD Model Tax Convention (2025), paragraphs 44.10 through 44.18, and the examples in paragraph 44.21. See also K. Cejie, “Let’s End the Fuss about Home Offices as Permanent Establishments for Cross-Border Teleworking,” Bulletin for International Taxation, November 2024, pp. 426–435.
For job categorization and risk profiles: MBB 2026/25, paras. 3.2 and 4; Vedder & Babuni, WFR 2023/90, para. 3.
Agreement between the competent authorities of the Netherlands and Belgium on the interpretation of Article 5 of the tax treaty (permanent establishment) with respect to employees working from home, Government Gazette 2023, 33856, December 8, 2023; C. Douven, NLF 2024/0018.
Douven, NLF 2024/0018, under “Greater interest on the part of the employer” and “Working from home leads (in exceptional cases) to a permanent work arrangement.”
BMF Letter dated February 5, 2024, IV D 1 - S 0062/23/10003:001 (Section 12 of the German Fiscal Code (AO)), in part in response to the Federal Fiscal Court (BFH) decision of February 22, 2023, X R 8/21; BMF Letter of December 24, 2025, IV B 2 - S 1301/01508/004/038. See Jacobs & Kühl, MBB 2025/11, para. 4.2.3, and MBB 2026/25, para. 7.
See Editorial Board, “Developments in Cross-Border Telework,” NDFR 2025/925, and Editorial Board, “Current Status of Cross-Border Telework,” NDFR 2026/421. See also NLF 2025/1306.
Framework Agreement on the Application of Article 16(1) of Regulation (EC) No. 883/2004 in Cases of Habitual Cross-Border Telework, effective as of July 1, 2023; Art. 13 of Regulation 883/2004 in conjunction with Art. 14 of Regulation 987/2009.
Art. 19 of Regulation 987/2009. For information on the Dutch approach and the interplay between taxation and social security, see NTFR 2022/82 and NDFR 2025/925.
Art. 15, para. 1, of the OECD Model Tax Convention and Art. 15, para. 1, of the Netherlands-Belgium Treaty. See Douven, NLF 2024/0018.
Art. 7.8 of the Income Tax Act 2001. For more information on qualifying foreign tax liability in cross-border worker situations, see also the explanatory notes in the 2026 Payroll Taxes Handbook, Chapter 19.
Commentary on Article 5 of the OECD Model Tax Convention (2025), para. 44.1, and Example A in para. 44.21.
In the case of a permanent establishment, profit allocation is determined in accordance with Article 7 of the OECD Model Tax Convention and the Authorized OECD Approach. The transfer pricing aspects are not discussed further in this article.
Supreme Court, October 14, 2022, ECLI:NL:HR:2022:1436, paras. 3.2.1 through 3.2.4; Articles 31 and 32 of the Vienna Convention on the Law of Treaties.
Similarly, Bane, NLF-W 2026/1, para. 3.1.3; Adriaans, NLF 2025/2607; MBB 2026/25, para. 8; Harteveld & Pötgens, WFR 2026/112, para. 4.
See Vedder & Babuni, WFR 2023/90, para. 4, and Douven, NLF 2024/0018.
We take the reference to the kitchen table from Vedder & Babuni, WFR 2023/90.
