A casual reading of the news these days leaves the impression that the United States mostly imports manufactures and exports food and raw materials. Ironically, this leaves out the category of international commerce that looms large in U.S. income statistics and drives U.S. economic growth: services. And this neglect is no small matter. It distracts from the reality that barriers to trade in services are far more problematic than barriers in goods.

First some background. Services have long played second fiddle to manufactures and agriculture in international trade negotiations. In part, that’s because of the common but long-outdated notion that almost all services are like hair stylists: living life as a blonde may be a lot cheaper in Beijing than Chicago, but there’s no practical way to stop by for a touch-up if you reside in Illinois.

To be sure, the market for many services is limited by geography – think restaurants, live concerts and hotel rooms. But even these services can be enjoyed by visitors from abroad, which amounts to an export by another name. In fact, in 2024 foreign tourists spent a whopping $215 billion in the United States, while American tourists spent about $177 billion outside the country.

Besides the obvious example of tourism, though, an array of business and consumer services are traded across borders in the internet age or supplied by service firms through their investments abroad. These include financial, accounting and legal services, higher education, all manner of computer services, movies, TV and musical entertainment, medical and engineering consultants and a plethora of niche services.

The figure to the right shows that two-way U.S. services trade has increased steadily since 2015, except for the entirely understandable dip in 2020 due to Covid-19. Over the period, service exports increased 44 percent to reach $1.1 trillion while imports rose 63 percent to exceed $800 billion. Note that the U.S. has enjoyed an annual services trade surplus for over two decades, making all too clear America’s competitive strength – indeed, dominance – in a wide range of services.

The figures on page 15 refine the picture, showing U.S. service exports and imports broken down by categories. Not surprisingly, the top three export categories in 2024 are travel, financial services and the diverse catchall “other business services.” That same year, the top three import categories were travel, transport (all those container ships) and … other business services

Nor is it surprising that digital tech – telecommunications, computer and information services – led export growth with an expansion of 90 percent in the decade. Likewise, most service import categories experienced robust growth, with personal, cultural and recreational services more than tripling. We Americans do enjoy a good time abroad.

When you picture the Great American Job Machine, images of workers beavering away on production lines at GM, U.S. Steel and Goodyear probably still come to mind. But today, the top five firms in terms of employment are Walmart, IBM, United Parcel Service, Target and Kroger. Not a single goods manufacturer makes the top 10.

No shock, then, that service industries on average accounted for around 86 percent of total U.S. non-farm employment during the period 2015 to 2024. The figure on page 16 shows the labor force divided into service-providing and goods-producing industries. Apart from the decline observed at the beginning of 2020, employment growth in service industries has been moderate but positive, increasing from 121 million to 137 million between 2015 and 2024. Note that while the share of workers in goods production has roughly held steady, value-added in the sector has risen by 12 percent (in spite of the Covid-19 dip).

Domestic U.S. Services Trade

In pioneering analysis, J. Bradford Jensen at the Peterson Institute devised a novel technique to measure services trade between U.S. metropolitan areas. Assuming that the consumption of different services commands almost the same share of income from one region to another, he examined detailed employment statistics for several service industries. When local production, measured by employment, exceeds local consumption, the excess is assumed to be exported.

Building on this insight, Jensen and colleague Antoine Gervais did a deep dive into internal U.S. commerce to determine the “tradability” of various sectors by applying a trade cost statistic. They found that 78 percent of industry value-added was essentially non-tradable between U.S. regions, while 22 percent was tradable. Some 12.7 percent of tradable value-added was produced by manufacturing industries and 9.7 percent by service industries. In other words, value-added by tradable services in domestic commerce amounted to about three-fourths of value added by tradable manufactures.

What’s this got to do with foreign trade? In 2024, U.S. exports of services totaled just $1,108 billion, 68 percent of exports of manufactures ($1,108 billion versus $1,638 billion). Put it another way: if U.S. services exports were the same proportion to value added in manufactured exports, they would have been $100 billion greater. High barriers to service exports largely explain the shortfall.

Actually, the shortfall in services trade is even larger when viewed on a global scale. In 2024, world exports of services amounted to $8.6 trillion, while world manufactures exports were $15.9 trillion. If the Gervais and Jensen calculation of tradability for services and manufactures can be applied globally, services exports should have been around three-fourths the size of manufactures exports. Instead, world service exports were only 54 percent of world manufactures exports. High barriers at borders go a long way to explaining the shortfall.

Stris, Aves and Other Barriers

Tariffs on services were never contemplated by American policymakers before Trump proposed a 100 percent movie tariff in May 2025. Years earlier, in the same nationalistic spirit, European countries designed digital services taxes as a way to extract revenue from U.S. tech giants.

But centuries before these mercantilist innovations, ingenious protectionists devised multiple ways of excluding or limiting foreign service suppliers. The OECD, which includes most high-income economies, catalogued a long list of barriers. For example:

• Foreign business ownership may be prohibited or allowed only up to a minority share.
• The sourcing of goods for government projects may be limited to domestic firms (e.g., Buy America).
• Job training requirements may exclude foreigners (common in medicine and law).
• Regulators may ban or apply special oversight conditions on foreign suppliers of services like telecommunications or banking.
• Maritime and civil aviation rules often restrict foreign carriers from transporting goods or passengers between domestic destinations (think New York to New Orleans).
• Private courier services like UPS and FedEx are often restricted in their scope of operations with the goal of reducing competition with government postal services.

Adding to the thicket of restrictions, while members of the World Trade Organization are countries, licensing and permitting rules for services are often set by sub-federal states or provinces not directly controlled by central governments.

Not only has the OECD catalogued the barriers, it has made a specialty of measuring them. First published in 2014 and now updated annually, the Services Trade Restrictiveness Index quantifies the burden that discriminatory regulations place on services trade.

Along with aggregate indexes, the OECD publishes STRI values for 22 major sectors in 51 countries, with index values ranging from zero (no barriers) to one (complete prohibition). And the calculations do contain surprises. Contrary to conventional wisdom, Japan is the least restrictive, with an aggregate index of 0.11; the United Kingdom is just behind with an STRI of 0.12. At the other extreme, the Philippines weighs in with an STRI of 0.45. In North America, the United States is 0.17, Canada is 0.20 and Mexico is 0.22.

Though obscured in the averages, the individual STRI sector indexes are all over the place. Perhaps not surprising, logistic and distribution services for handling merchandise tend on average to be burdened with few discriminatory restrictions. Both motion pictures and computer services have average indexes of just 0.20. At the other extreme are air transport (0.40), reflecting controls on ownership, routes and landing rights, and legal services (0.34), reflecting protectionist licensing requirements.

Researchers have combined STRI values with measured trade flows, distance between countries and their economic size in “gravity models” to estimate the tariff equivalents (AVE) of service trade barriers. Last year Andrew Bridger of the University of Edinburgh estimated tariff-equivalent rates for seven sectors and 50 countries. The results are startling, with some AVEs exceeding 100 percent. In fact, for finance, the median estimated AVE for 50 countries is 168 percent; for business services, 44 percent. In that latter category, Chile was at the low end of the 50 countries with an AVE of just 16 percent, while India was at the high end at 148 percent. The UK’s AVE was 31 percent, while the U.S. clocked in at 26 percent.

Gatt and Gats

In 1947, 23 of the World War II victors led by the U.S. and UK formed the General Agreement on Tariffs and Trade. The core objectives were to slash high tariffs inherited from the Great Depression (think Smoot-Hawley) and to codify rules for the conduct of international commerce. One key rule was the most-favored-nation (MFN) principle: all trading partners of each member-state are entitled to the lowest line-item tariff imposed by that member. Another was the national treatment principle: imports are entitled to the same tax and regulation treatment as domestic competitors.

But the scope was relatively narrow: tariff cuts and commercial rules in the GATT were tailored to trade in manufactures, agriculture and mining. Indeed, the only important service covered by GATT was freedom of transit, and this as an auxiliary to merchandise trade.

Not until the so-called Uruguay Round of Multilateral Trade Negotiations (1986-1994), which transformed the GATT into the World Trade Organization, were services nominally granted an equal place at the table. The enabling treaty, the General Agreement on Trade in Services (GATS) was only established in 1995.

To be sure, long before GATS, commerce in certain services was coordinated by a host of special-purpose international treaties and agreements including the International Telecommunication Union (1865), the Universal Postal Union (1874), the Bank of International Settlements (1930), the International Maritime Organization (1948) and the Arrangement on Officially Supported Export Credits (1978). But unlike these special-purpose pacts, the GATS sought to cover all services with uniform rules, preparing the ground for future liberalization. And again, in contrast to the special-purpose pacts, GATS applied both to all 76 incumbent WTO members in 1995 and to new members that subsequently joined.

As with GATT, GATS embraced two basic principles: most-favored-nation and national treatment. But GATS did not require any liberalization of existing practices by member countries. Members could, if they chose, simply announce deviations from MFN (and there were many), while national treatment only applied when members positively accepted the obligation, sector by sector. And in almost all cases, members only accepted national treatment in sectors where that had been the country’s practice before the 1995 services agreement.

Along with codifying 12 broad service sectors and 150 sub-sectors, GATS identified four “modes” of service delivery:

• When a person travels to another country and consumes goods and services there (think tourism)
• When a person or firm buys services supplied by a foreign country (often online, but also from foreign consultants)
• When a firm invests in a service enterprise abroad (foreign investment by, say JPMorgan Chase or Yale University)
• The movement of individuals (think Taylor Swift or the Berlin Philharmonic)

The hope of GATS’s cheerleaders was that all members would apply MFN and national treatment to all modes of delivery in all sectors. While progress was made in this direction in specialized agreements on financial services and basic telecommunications, that proved the end of the line. Those negotiations, it seems, exhausted the willingness of most WTO members to liberalize their service sectors. But the resulting deadlock did prompt a subset of liberalizing spirits among WTO members to try a less ambitious approach, a “plurilateral” agreement limited to a coalition of the willing.

Trade in Services Agreement

During President Obama’s second term, 23 countries led by the United States tried to negotiate a plurilateral Trade in Services Agreement (TiSA). The 23 countries, mostly advanced economies, called themselves “Really Good Friends of Services.” Brazil, India and Russia were not members, and the U.S. blocked China from joining. The U.S. ambassador to the WTO, Michael Punke (who also happens to be the author of the bestselling novel The Revenant!) led the talks. But unlike Revenant’s protagonist Hugh Glass, who wrestled with a grizzly bear, TiSA did not survive.

Formal negotiations, outlined in a Peterson Institute Policy Brief, began in March 2013. After 21 rounds, negotiations were suspended in November 2016. The negotiators had aspired to incorporate TiSA in the WTO framework, although commitments would be made by only the 23 “friends” out of the 164 WTO member countries. Joining TiSA to the WTO would have required giving TiSA benefits to all WTO members, even though only the 23 “friends” were obligated to reciprocate. This proved a major sticking point.

The TiSA draft contemplated national treatment commitments in computer, telecommunications, information technology and general business services. It did not cover entertainment or services funded by public authorities, such as health and education.

The U.S. sought complete freedom for cross-border data flows, but the EU objected on privacy grounds.

US-UK Deal: Missed Opportunity

Fast forward almost a decade. Announcing the hurry-up U.S.-UK trade deal on May 9, 2025, the U.S. Trade Representative stressed new openings for U.S. beef and ethanol exports and preferential UK access to the U.S. auto market. Concessions on digital services and construction received passing notice, but merchandise was the centerpiece. This suited the political needs of both President Trump and Prime Minister Starmer. But services, not merchandise, are the centerpiece of the U.S.-UK trade and investment relationship, and service barriers deserved more attention.

In 2024, two-way services trade was $240 billion, with the U.S. exporting some $70 billion to the UK, and the UK exporting $170 billion to the U.S. By contrast, two-way merchandise trade was just $148 billion. But the focus on goods was understandable in light of Trump’s focus on the merchandise trade.

While the UK had a very low average STRI (just 0.12) in 2024, three of its relatively restrictive service sectors, motion pictures (0.15), accounting services (0.25) and air transport (0.20) are important to American business interests. A deal including liberalization would have added significant value to the agreement.

U.S. restrictions on services trade are somewhat greater than in the UK, but still low in aggregate with an average STRI of 0.17 in 2024. However, four sectors are highly protected: air transport (0.52), insurance (0.28), water transport (0.36) and courier services (0.35). All four should be of interest to British business interests (not to mention U.S. consumers), and UK negotiators should investigate details of the U.S. barriers.

The Digital Revolution

In 1998 at the dawn of the internet age, the WTO agreed to a two-year e-commerce moratorium on protectionist initiatives. The moratorium, renewed every two years since its inception, prohibits customs duties on electronic transmissions of data, books, articles, sound and visual material. India, Indonesia and South Africa strongly opposed the 2024 renewal (expiring in March 2026). The precedent of Trump’s 100 percent movie tariff initiative (even if blocked by Congress or the courts), along with opposition from emerging market countries, may soon spell the end of the e-commerce moratorium and the initiation of customs duties on electronic transmissions.

Meanwhile, for three decades the digital revolution has enabled global services trade to flourish. Amazon is the prime example, with platforms for retail transactions, videos and music. The WTO calculates that digitally delivered services amounted to $4.6 trillion in 2024, more than half of the $8.6 trillion in commercial service exports. Digitally delivered service exports increased by almost a factor of four from 2005 and 2022, while other service exports only doubled. And that may only be the beginning: Jupiter Research forecasts that cross-border e-commerce sales from business to consumers alone will reach $3.3 trillion in 2028, more than double the figure in 2023.

* * *

It is the best of times and worst of times, then, in global trade in services. While key nations are unwilling to move forward on liberalizing traditional services like shipping and finance, the combination of digital delivery and artificial intelligence promises a bright future for the electronic services.

The unanswered question, then, is whether the combination of nationalism and protectionism – the anti-globalization forces of the world today – will block that promise. Hence from our perspective, the best hope is benign neglect.