Why the USA and Serbia Need a Tax Treaty?

Why the USA and Serbia Need a Tax Treaty

Introduction

According to the Serbian government, the first U.S.–Serbia Strategic Dialogue opened on July 17, 2026, establishing a regular framework for cooperation across foreign policy, security, energy, trade, investment, science, education, innovation, and advanced technologies. This is critical because tax and social-security coordination do not sit neatly inside a single ministry, and as such any serious effort would require several agencies to work from the same bilateral agenda.

The two countries nevertheless have neither a comprehensive income-tax treaty nor a social-security totalization agreement: Serbia is absent from the IRS list of current U.S. income-tax-treaty partners and the SSA list of countries with totalization agreements. Commerce, of course, continues without them, but cross-border payments remain more dependent on domestic tax rules, while mobile workers have no bilateral mechanism for coordinating social-security coverage. The absence of an agreement is therefore not an impassable barrier, but could create frictions by determining which domestic rules companies and workers must navigate when activity crosses the border.

The relationship is consequential enough for this gap to need addressing. USTR reports about $1.2 billion in U.S.–Serbia goods trade in 2025 and about $1.2 billion in services trade in 2024. The services side is particularly relevant because many of the issues addressed by tax treaties arise through licensing, financing, professional activity, and the allocation of income rather than through the physical movement of goods.

The U.S. Commercial Service calls the United States Serbia’s largest ICT trading partner and identifies Oracle, Rivian, Microsoft, NCR Atleos, NCR Voyix, Cisco, and IBM among major U.S. companies or affiliates with Serbian operations. Therefore, ICT presents itself already as one of the most robust areas in which the two economies interact. Combining services, intellectual property, investment, and related-party transactions, Serbian government data put ICT-services exports to all markets at €4.552 billion in 2025, 10 percent above 2024. That growth increases the practical importance of rules that distinguish ordinary cross-border services from royalties, taxable business presence, and related-party income.

Capital-market links have also deepened. In May 2026, Telekom Srbija announced a €1.95 billion multi-tranche international bond issuance that attracted $13.87 billion in orders from roughly 300 institutional investors. The transaction does not itself make the case for a bilateral treaty, but it illustrates the extent to which Serbian issuers now operate in markets where the tax treatment of cross-border interest and investor status can matter. Does domestic law alone remain enough for a relationship that now includes technology, intellectual property, international finance, investment, and professional mobility? That is the more useful question, not whether a treaty sounds desirable in the abstract, but whether identifiable bilateral problems are significant enough to justify negotiating one.

Policy Recommendations 

1. Totalization scoping. The two governments should designate technical teams to determine whether Serbia’s pension and disability insurance system can support an agreement under Section 233 of the Social Security Act, including coverage categories, certificates of coverage, benefit coordination, and administration.

2. Treasury–Ministry of Finance consultations. Serbia should use the Strategic Dialogue to present the U.S. Treasury Office of Tax Policy with a documented inventory of bilateral tax frictions rather than treating treaty negotiations as automatic.

3. A bilateral technical working group. A standing group under the Strategic Dialogue could compile data, compare legal frameworks, and maintain continuity across changes in political leadership without committing either government to negotiations.

4. A defined Serbian negotiating position. Before formal talks, Serbia should identify its preferred treatment of dividends, interest, royalties, permanent establishments, capital gains, dispute resolution, information exchange, and anti-abuse rules, including the tradeoff between source-country revenue and lower cross-border tax friction.

The Legal Baseline

U.S.–Yugoslav Treaty Negotiations

The historical record shows repeated, unsuccessful attempts to negotiate a U.S.–Yugoslav income-tax treaty. The Treasury’s fiscal 1976 annual report records negotiations and technical discussions with Yugoslavia and a November 1978 Federal Register notice later classified the talks as inactive, however, the effort later resumed. A 1981 Federal Register notice announced discussions in Washington for March 10–13, and the Treasury’s 1987 release index records another April 10 announcement that the United States and Yugoslavia would negotiate an income-tax treaty. The 1987 entry is especially important because it pushes the record beyond the early-1980s talks and shows that the issue resurfaced again before Yugoslavia’s dissolution.

However positive the momentum was, no convention was ultimately concluded. In 1999 testimony, the Treasury’s International Tax Counsel described Slovenia as the first country in the former Yugoslav area with which the United States would have concluded a tax treaty. Any modern U.S.–Serbia effort would therefore draw on a history of prior negotiations without inheriting an agreement that could simply be revived.

The Western Balkan Trend Vs. Serbia’s Current Tax Treaty Network 

Beyond Slovenia, the Treasury’s treaty records show that Croatia is far along in the process though not quite at active status, signing its first comprehensive U.S. income-tax treaty in December 2022 and a protocol in April 2026; Treaty Doc. 119-2 records their September 14, 2026 transmission to the Senate. Unlike its regional neighbors, Serbia has been maintaining an extensive treaty network elsewhere. The OECD's sixth peer-review report on treaty shopping identifies 64 tax agreements in force. Under Article 40 of Serbia's Corporate Income Tax Law, specified payments by resident legal entities to nonresident legal entities are generally subject to a 20 percent withholding tax unless an applicable tax treaty provides otherwise. These payments include dividends, royalties, interest, leasing income, and certain consulting and professional services. Article 40a conditions treaty relief on proof of residence in a treaty-partner jurisdiction and beneficial ownership of the income. Consequently, otherwise comparable cross-border payments may face different withholding-tax treatment depending on the recipient's tax residence, legal status, and eligibility for treaty benefits.

Figure 1. Serbian source-country withholding: domestic law and two treaty benchmarks. The 2016 U.S. Model is a negotiating baseline, not a Serbia forecast; its 0% interest and royalty values represent the general Model rule and are subject to exceptions. Sources: Serbian Corporate Income Tax Law, Art. 40; China–Serbia DTT; 2016 U.S. Model Income Tax Convention, Arts. 10–12.

What an Income-Tax Treaty Would Change

The 2016 U.S. Model illustrates how an income-tax treaty allocates taxing rights rather than simply lowering every rate. Its provisions address permanent establishments, withholding, a Mutual Agreement Procedure (MAP), and information exchange. For Serbia, those benefits would have to be weighed against the source-country revenue surrendered through lower negotiated rates. A treaty does not make cross-border income tax-free. It determines which country may tax particular income, when source-country tax is limited, and how the two systems interact when both claim taxing rights.

Some friction is already mitigated unilaterally. IRS guidance says U.S. taxpayers may be able to claim a foreign tax credit for qualifying foreign income taxes, while Treasury’s FATCA materials record Serbia’s Model 1 intergovernmental agreement with the United States as in force since January 8, 2020. Neither mechanism substitutes for bilateral rules governing source taxation, permanent establishments, MAP, or treaty eligibility. The foreign tax credit can reduce some double-tax exposure after tax has been imposed, but it cannot create the negotiated limits and procedures that a treaty would establish in advance.

Furthermore, Blonigen and Davies found no systematic evidence that bilateral tax treaties increased U.S. FDI activity, while an IMF study of 41 Sub-Saharan African economies found that treaties with investment hubs were not associated with additional investment in its sample and could involve source-country revenue losses. Neither result is Serbia-specific, though together they serve to caution against treating signature itself as an economic outcome. This shows that a treaty cannot generate investment on its own, and should be perceived more as a tool to assist and strengthen ties by using it towards specific tax and administrative problems.

Where the Current Gaps Matter Most

ICT, Intellectual Property, and Related-Party Transactions

Technology is central to the bilateral services relationship. IRS sourcing guidance generally looks to where personal services are performed, so work performed entirely in Serbia does not become U.S.-source merely because the customer is American. The more treaty-sensitive issues arise when people, taxable business presence, intellectual property, or related entities cross borders. A Serbian developer working from Serbia for an American customer is a different tax problem from an employee temporarily working in the United States, a company with a U.S. permanent establishment, or a payment that is characterized as a royalty rather than a service fee.

Transfer-pricing disputes are a useful example. Either tax administration can adjust intercompany charges for engineering, software development, management services, financing, or intellectual property. A treaty would not eliminate that uncertainty, but the 2016 U.S. Model’s Mutual Agreement Procedure illustrates the formal channel available to competent authorities to address double taxation caused by inconsistent adjustments. If one tax authority increases the income allocated to an entity and the other does not make a corresponding adjustment, the same income can effectively be taxed twice. MAP gives the two competent authorities a mechanism for addressing that problem even though it does not guarantee a particular outcome.

Royalties create a separate source-tax issue. Without treaty relief, qualifying Serbian-source royalty payments to a U.S.-resident legal entity can fall within Serbia’s domestic withholding regime. The 2016 U.S. Model generally assigns royalty taxation to the residence country, although actual U.S. treaties can preserve source-country taxation and special rules can apply. The negotiated treatment would matter for software licenses, gaming IP, patents, and cloud or SaaS arrangements when the payment is characterized as a royalty. The classification itself is relevant because a payment for a service and a payment for the right to use intellectual property can produce different source-tax results.

Acquisitions, Holding Structures, and Capital Markets

U.S. buyers have acquired businesses rooted in Serbia’s technology sector. Take-Two’s SEC filing, for example, describes Nordeus Limited as an Irish holding company of a Belgrade-based game developer. That public disclosure establishes part of the corporate structure, but not whether treaty considerations drove the transaction’s tax structuring. The distinction comes into play when the existence of an intermediate holding company does not, by itself, establish why that jurisdiction was chosen.

More generally, investors between countries without a bilateral treaty may use treaty-covered holding jurisdictions when commercial and legal considerations support them. Such structures remain subject to beneficial-ownership, substance, domestic anti-abuse, and treaty anti-shopping rules. A direct U.S.–Serbia treaty could make treaty access less important in otherwise bilateral structures, while leaving financing, governance, regulatory, and other reasons for intermediate entities intact. In other words, a treaty could change one reason for using an intermediate jurisdiction without eliminating the broader corporate reasons for doing so.

The same basic issue applies to financing. Telekom Srbija’s 2026 issuance demonstrates Serbian access to global institutional capital, but the tax treatment of any investor depends on the security, investor status, payment terms, exemptions, and structure. Serbia’s domestic law places interest among the payment categories potentially subject to withholding, making it one area in which treaty terms can materially affect cross-border financing. This is precisely why the eventual treatment of interest would matter in negotiations, as the answer depends on the negotiated rule, not simply on the size of the capital-market transaction.

Phase One: Social-Security Totalization

A totalization agreement is narrower than an income-tax treaty and addresses a different problem. The Social Security Administration (SSA) describes totalization agreements as mechanisms that generally prevent dual social-security taxation in covered employment situations and allow qualifying workers to combine periods of coverage accumulated in two systems. The practical case is easiest to see with a temporary assignment: without coordination, an employer and employee can face questions over which system applies, while a worker with a career split across countries may need a way to combine qualifying coverage periods.

Section 233 of the Social Security Act authorizes agreements with countries that have qualifying social-security systems. Serbia’s Pension and Disability Insurance Fund covers compulsory-insured employees, self-employed workers, and farmers and lists social-insurance agreements with 35 countries. Those justify a technical compatibility review; they do not predetermine the U.S. legal or administrative conclusion. The first step would subsequently be technical, comparing the systems, determining which categories of employment could be coordinated, and identifying which administrative procedures would be required.

The U.S. process also differs from an Article II treaty in that a totalization agreement does not require two-thirds Senate advice and consent. Under Section 233, it is transmitted to Congress and must remain there for a statutory 60-session-day review period before it can enter into force. The Social Security Administration leads the U.S. technical process; on the Serbian side, the Ministry of Labour, Employment, Veteran and Social Affairs would likely coordinate with the PIO Fund and other implementing institutions. That narrower statutory route is one reason totalization can be evaluated separately from the broader tax-treaty question.

The SSA’s current agreements list shows 31 U.S. totalization agreements in force. Its Romania guidance identifies September 1, 2026 as the effective date for the newest agreement. Romania is useful as a recent comparison because it shows that even a signed agreement can take years to reach implementation. For reference, the SSA status table shows that Poland took about 11 months from signature to entry into force; the Czech Republic, Slovak Republic, and Hungary fell roughly in the 16–19 month range; and Slovenia took about 24.5 months; but Romania took about 41 months. So, totalization is best treated as the narrower first track because its technical feasibility can be tested independently of the Treasury’s decision on an income-tax treaty. These examples should not be converted into a forecast for Serbia but show a range of outcomes and, most importantly, that the period after signature is not fixed.

Phase Two: Income-Tax Treaty

Treasury Prioritization and the Negotiating Baseline

The Treasury’s current public treaty materials do not provide a ranked queue of prospective partners, while historical Treasury testimony describes the complexity of reconciling two tax systems and the need for a treaty to justify its negotiated compromises. Serbia would strengthen its case by documenting concrete bilateral problems rather than assuming that commercial ties alone will generate negotiations. The strongest submission to the Treasury would be a record of actual frictions, e.g. recurring withholding issues, unresolved classification questions, business concerns, and administrative problems that cannot be handled adequately under existing law.

The 2016 U.S. Model Income Tax Convention remains the clearest public baseline for the U.S. negotiating position. It addresses dividends, interest, royalties, capital gains, permanent establishments, information exchange, MAP, and detailed limitation-on-benefits rules. Actual treaties depart from the Model when domestic laws and negotiating priorities require compromise, so its rates should be read as a starting point rather than a Serbia forecast. For that reason, the Model is most useful for identifying the subjects that would have to be negotiated, not for predicting the final numbers.

Likely Negotiating Issues

The core bargaining questions would center on source-country taxing rights and anti-abuse protections. The 2016 U.S. Model generally limits source taxation of interest and royalties and includes detailed anti-abuse and information-exchange provisions; Serbia may seek to preserve more source taxation of interest, royalties, services, or business profits. Permanent-establishment and capital-gains provisions would matter for projects, acquisitions, restructurings, and other cross-border operations. A lower withholding rate can reduce tax friction, but it also gives up source-country revenue, while broader source taxing rights can preserve revenue at the cost of leaving more tax in the transaction.

A U.S.–Serbia convention would also be expected to contain a detailed limitation-on-benefits article consistent with current U.S. model policy. Serbia already operates its own anti-treaty-shopping framework: the OECD peer review reports 64 tax agreements in force, 43 of which comply with the Action 6 minimum standard, and records Serbia’s 2017 signature and 2018 ratification of the Multilateral Instrument as well as use of treaty preambles and a principal-purpose test. Negotiators would not be starting from the question of whether anti-abuse rules are needed, but the harder issue would be how the U.S. limitation-on-benefits approach and Serbia’s existing framework fit together.

Negotiation and Senate consideration are separate sources of delay. A multi-year planning horizon is realistic while a fixed 36–48 month prediction is not. Even a successful negotiation would only move the process into a second institutional stage in the United States.

Using the Strategic Dialogue

The Strategic Dialogue provides an institutional channel for preparatory work without committing either government to a treaty. On September 17, Serbia’s foreign minister said the format now includes annual cooperation plans for ministries, agencies, and other institutions, with defined action plans in several fields. That statement does not create a tax or social-security workstream, but it supports using the Dialogue for agency-level follow-through while the Ministry of Finance, the Treasury, SSA, and Serbian social-insurance institutions retain their formal legal roles. The value of the Dialogue is therefore coordination, keeping the issue moving while the agencies with actual statutory authority do the technical work.

A technical working group could begin with four tasks: a) estimate the number and types of workers potentially exposed to dual social-security coverage; b) identify the principal bilateral payment categories subject to withholding; c) compare Serbian law with the U.S. Model and relevant Serbian treaty precedents; and d) reconstruct the unresolved issues from the earlier U.S.–Yugoslav negotiations. That record would help separate matters requiring a treaty from those that could be handled through administrative clarification or a separate totalization agreement, as well as give both governments something more useful than a general statement of interest, a common factual record from which they could decide whether formal negotiations are worth opening.

Political continuity is another reason to keep the first stage technical. Leadership changes can alter priorities; ministries can still preserve data, legal analysis, and draft positions for the next government to evaluate. Keeping the early work technical would reduce the extent to which the entire process depends on one political officeholder or one electoral timetable.

Risks and Constraints

Negotiating capacity and priority could be major constraints. Treasury testimony describes tax-treaty negotiations as time-consuming because of the multiple functions treaties perform and the need to mesh two tax systems. Totalization follows a different statutory path but still requires negotiation, administrative preparation, congressional review, and implementation. The practical constraint is not only political willingness, but also whether the responsible agencies have enough reason to devote scarce technical capacity to the work.

The substantive tradeoffs are equally important. Serbia may want to preserve source-country revenue, while the United States may seek provisions close to current U.S. model policy on anti-abuse and information exchange. Rates that are too high can leave little commercial gain and rules that are too permissive can create treaty-shopping concerns. The objective should be a treaty whose provisions solve identifiable bilateral problems, not an agreement pursued for its own sake. Before that though, Serbia would need to decide in advance where lower tax friction is worth giving up source-country revenue.

Conclusion

The bilateral economic relationship between the U.S. and Serbia is now broader. The significance of that change is not simply that the relationship is larger but that more of the relationship now runs through the kinds of payments, corporate structures, assignments, and disputes that bilateral tax and social-security agreements are designed to address.

The most workable sequence is to begin with totalization scoping, use the Strategic Dialogue to build the factual record for Treasury–Ministry of Finance consultations, and develop a Serbian negotiating mandate before formal tax-treaty talks. Negotiations should follow only if both governments conclude that the expected benefits justify the revenue, administrative, and anti-abuse tradeoffs. The immediate task is not to announce an agreement, but to establish whether the technical and commercial case is strong enough to negotiate one.