Imagine if Congress decreed that Chinese investors should pay lower taxes on income from U.S. assets than U.S. residents pay. Surely, one might think, that could not happen. Why would Congress encourage the passing of U.S. assets to Chinese control? And why would it choose to forgo the tax revenue when that did happen? And why, at a time when Democrats and Republicans are united in wanting to end the dumping of Chinese government–subsidized goods in the United States, would it want to add to those subsidies?
Yet Congress is, however unintentionally, doing all of these things. U.S. law does, in fact, subsidize foreign investment in the United States by imposing much lower tax rates on foreign investors than American ones. And that subsidy gives an advantage to imported goods over home goods, as dollars earned by foreigners can be more profitably invested in the United States than dollars earned domestically.
In the language of macroeconomics, the United States runs a large and persistent current-account deficit with China—meaning, roughly, that it imports far more goods and services from China than it exports. As a necessary mirror image of that deficit, it also runs a capital-account surplus with China—meaning that it imports more capital than it exports. Although it is widely understood that the United States would necessarily import less Chinese capital if it imported fewer Chinese goods, it is generally not understood that the reverse is equally true. So if the U.S. tax subsidy for the import of foreign capital were eliminated, Chinese investors would have less motivation to outbid Americans for U.S. assets and, by extension, less incentive to dump goods in this country in return for dollars.
NECESSARY STEPS
China spends nearly five percent of its gross domestic product subsidizing domestic industrial firms. The United States, by contrast, spends roughly 0.4 percent of GDP on such subsidies. Nearly all of China’s 5,300 publicly listed companies disclose financial subsidies from the state, and these sums are often huge. The Chinese battery maker CATL, for example, received $790 million in state subsidies last year. Such subsidies, which are directed primarily at exporters, are more than enough to give Chinese firms an artificial and, often, insurmountable competitive advantage internationally.
For many decades now, the United States has imported far more goods—and therefore more capital—from China than it has exported. In 2024, China’s goods-trade surplus with the United States amounted to $295.4 billion. This recurring annual surplus is larger than it would be if the United States ceased supporting it with huge tax breaks for Chinese investors.
Whereas trade deficits are not inherently bad for economic growth, any more than surpluses are inherently good, they harm more efficient U.S. firms and more productive U.S. workers when fueled by foreign subsidies and other state supports. This mattered less when China, sporting a much smaller economy, joined the World Trade Organization in 2001, but it clearly matters now that its economy is 1,400 percent larger. Today, if the Chinese government seeks to dominate a given global industry, such as electric vehicles, it can do so through state support.
To counter China’s mercantilist trade policy, then, Congress should revise U.S. tax law. Amending sections of the Internal Revenue Code and withdrawing from the U.S.-China income tax treaty would result in significant changes in U.S. international tax policy, but all the necessary steps are well within Congress’s power. Such an approach would—unlike the Trump-Biden tariffs, under which the total annual current-account deficit has soared to $1 trillion—actually raise revenues from abroad and encourage more American investment at home.
FIX THE TAX CODE
U.S. tax law, like that of most countries, draws a sharp distinction between the operations of a foreign-owned business in the United States—foreign direct investment, or FDI—and passive foreign investment in U.S. securities. If a foreign company, say a Swiss bank, decides to open operations in New York, income from that business would be taxed by the United States as if it were a U.S. bank that earned it. When it comes to taxation of active business income, it does not matter that the foreign bank is organized and headquartered in Switzerland. Rather, it is the country where business income arises—the United States—that has the primary taxing claim. If Switzerland wants to tax its bank’s U.S.-source income as well, it is of course free to do so. In order to avoid taxing the same income twice, however, Switzerland would typically reduce its tax claim on that income by the amount of the tax already paid to the United States.
That the United States has a primary claim on business income generated within its borders makes good sense, since American workers and consumers, as well as U.S. legal and physical infrastructure, are responsible for the creation of that income. If Washington ceded taxing authority to the business’s home country, a company from a low-tax country would be able to open operations in the United States and trample U.S. competitors simply because it was subject to a lower tax rate.
Different rules apply when a foreign individual or entity decides to buy stocks or bonds of a U.S. company. As long as the investment is small compared with the company’s overall size—a “portfolio investment,” in financial terms—it has no effect on the company’s operations or management. One percent of Apple stock, for example, is worth over $35 billion. But a one percent shareholder would have no ability to shape the decisions of Apple’s board of directors or the priorities of Apple’s CEO. Moreover, if this investor is a foreigner, the investor’s connection to the United States is tenuous—in contrast with a company that opens a business in the country. Granting the investor’s home country a primary taxing claim therefore is a sensible approach. It further allows that country to tax the investor’s income using a progressive scale, if it so chooses. In contrast, the United States—the country of the passive income’s source—cannot do the same, as it has no knowledge of the foreign investor’s total income.
Still, Washington does tax some portfolio investments by foreigners. Specifically, it imposes a maximum withholding tax of 30 percent on interest and dividends received by foreigners from U.S. sources. It can do so successfully because the withholding agents tasked with enforcing U.S. tax rules are always U.S.-based institutions. For example, if a Hungarian individual holds a share of PepsiCo stock in an account at Goldman Sachs, and PepsiCo pays a $10 dividend on that share, Goldman would deposit $7 into the individual’s account and remit $3 to the Internal Revenue Service. And if that individual’s account is not at Goldman but at UBS in Switzerland, then PepsiCo itself would be responsible for sending $3 out of the $10 dividend to the IRS, remitting only $7 to the UBS account. In contrast, the United States cannot easily impose taxes on capital gains from selling U.S. securities, since foreigners can undertake sales wholly outside U.S. jurisdiction. So, for example, when a Hungarian individual sells PepsiCo shares held in a UBS account in Switzerland to a German individual, the IRS lacks the power to collect capital gains taxes on the sale. Having no power to collect these taxes, the United States exempts capital gains in such cases.
Congress should withdraw from the U.S.-China income tax treaty.
The U.S. withholding tax is not primarily a means for Washington to extract revenue from foreigners, however, but a means of extracting more of it from Americans. Washington uses the substantial 30 percent tax rate as bargaining leverage with other countries. In bilateral tax treaties, it reduces the U.S. withholding rate considerably in return for lower rates on interest and dividends earned by Americans abroad. Those lower foreign rates allow Washington to tax more of that income while avoiding double taxation. The U.S. bilateral treaty with China, for example, reduces each country’s withholding tax on the other’s residents to a mere ten percent.
Yet even this low rate is reduced to zero in several important cases. The tax code exempts from tax all portfolio income of foreign governments, including their sovereign wealth funds. Additionally, in an effort to make U.S. corporate borrowing abroad easier and cheaper (something that was already happening via foreign-incorporated shell companies), Congress exempted bond interest from withholding tax altogether. Finally, since financial derivatives took off in the 1980s, the Treasury Department has also exempted payments from U.S. counterparties to foreign ones.
Owing to the combination of these exemptions, the IRS’s limited jurisdiction, and the logic underlying the proliferation of bilateral tax treaties, the unintended result is that the U.S. tax code favors investments by foreigners over those by Americans. For example, a U.S. resident receiving dividend income from the American technology company IBM would pay tax of at least 15 percent, whereas a Chinese sovereign wealth fund investing in IBM would pay no U.S. tax at all. A U.S. resident receiving interest on an IBM bond or a Treasury security would pay tax at his or her marginal rate, which may be 37 percent or more. A Chinese sovereign wealth fund would pay nothing. The upshot is that the after-tax returns for U.S. investments by Chinese entities well exceed those by U.S. investors—providing a significant incentive for Chinese entities to buy up U.S. assets from U.S. issuers and owners. Unsurprisingly, there is a massive imbalance in capital flows between China and the United States. Whereas American investors held $322 billion in Chinese portfolio assets at the beginning of 2024, Chinese investors held six times that figure—$1.87 trillion—in American portfolio assets.
Notably, the more attractive the after-tax returns on dollar investments are to Chinese entities, the more goods China will sell in the United States to earn dollars. Changing the tax code to make those investments less attractive would reduce the Chinese government’s incentive to subsidize the export of steel, batteries, electric vehicles, and other goods—exports that U.S. policymakers are struggling to contain.
CURB UNWANTED INVESTMENTS
Congress should reform U.S. tax law to make it less attractive for the Chinese government and Chinese businesses to accumulate U.S. dollars. Eliminating tax preferences for Chinese investments in U.S. securities would require several steps. The first would be to remove the present exemption that the Chinese government enjoys from taxation of portfolio investment income. The United States is an international outlier in providing a blanket exemption from such tax for foreign governments, and there would be nothing radical about removing it. As for the legal probity of targeting such a change at one country—China—this would hardly be unprecedented. The Internal Revenue Code already contains exclusions from some favorable tax provisions for countries that do things contrary to U.S. policy or interest, such as supporting terrorism (Iran, North Korea, Sudan, and Syria), or for countries that boycott Israel (such as Kuwait and Lebanon).China would simply be another state to lose favorable tax treatment—in this case, owing to its distortionary trade practices. And the United States would not be alone in targeting Chinese government entities. Canada, for example, long ago repealed its tax exemption for Chinese state-owned banks.
Senator Ron Wyden, a Democrat from Oregon and chair of the Senate’s tax-writing Finance Committee in the previous Congress, already introduced legislation in 2023 that would deny the tax-free benefit to the government of China and that of other countries meeting certain criteria. Yet this step would accomplish little without further reforms. Of the $1.87 trillion that China holds in U.S. portfolio securities, almost $1.4 trillion are debt obligations, most of which qualify for a separate U.S. tax rule exempting foreign investors from tax on most types of interest income. Congress should repeal this rule for any payments made, directly or indirectly, to any Chinese investor. It also needs to make sure that the old trick used to avoid this tax—the setting up of special shell companies—no longer works. That process is explained below.
The next step would be for Washington to withdraw from the U.S.-Chinese tax treaty, which lowers the withholding tax on most dividends and interest income to ten percent. To make a dent in the U.S.-Chinese trade imbalance, the withholding rate needs, at least, to revert to that specified in the Internal Revenue Code—30 percent.
Enacting rules to curb unwanted Chinese investments is necessary, but if China can disguise investments in ways that render them unrecognizable to U.S. withholding agents the rules will not work. This is why Wyden’s draft bill refers to U.S. securities held “directly or indirectly” by nonexempt foreign governments. The bill delegates to the Treasury the task of writing regulations “to prevent the avoidance of the [bill’s] purposes.” Treasury would need to innovate. That process is also explained below.
It is not difficult to imagine that China’s enormous sovereign wealth fund, China Investment Corporation, would react to passage of these reforms by seeking to disguise its holdings—as rich taxpayers around the world have been doing for decades. More specifically, CIC would likely set up a shell company—call it Shellco—in a taxpayer-friendly jurisdiction having a favorable income tax treaty with the United States. It would then have Shellco open a trading account with a financial institution having no U.S. presence—call it Offshore Bank. CIC would continue to invest as before, except that the nominal owner of U.S. securities would thereafter be Shellco.
This simple scheme would be remarkably difficult to defeat. Payments on U.S. securities, such as interest and dividends, come from U.S. payers and are handled by financial institutions subject to U.S. jurisdiction. It is easy for Congress to designate all such institutions as withholding agents and require them to demand evidence of beneficial ownership—that is, the identity of the ultimate owner, rather than an agent, nominee, or other intermediary—for each account that they credit with a payment on U.S. securities. In fact, U.S. tax rules already impose this requirement. The problem is that Offshore Bank would simply tell the U.S. withholding bank requesting information that the account is owned by Shellco. CIC’s beneficial ownership would remain hidden from U.S. authorities.
Enacting rules to curb unwanted Chinese investments is necessary.
If Offshore Bank were subject to U.S. jurisdiction, Congress could impose on it “know-your-customer” rules that would require it to look through Shellco (or a series of “Shellcos”) until it found a true beneficial owner. Alas, Offshore Bank is outside U.S. reach. It might appear that Congress’s only choices would be to accept Shellco as a legitimate owner or require 30 percent withholding on all U.S.-source payments to foreign financial institutions—unless they submitted to U.S. know-your-customer-type rules. Governments globally would view the latter choice as an extraordinary assertion of U.S. extraterritorial jurisdiction.
Fortunately, there is a third choice that would take advantage of information already collected by financial institutions around the world: joining the Common Reporting Standard. In 2014, the Organization for Economic Cooperation and Development created the CRS, a multilateral, cooperative approach to the exchange of information about beneficial owners of financial accounts. Over 100 countries have since joined the CRS. The OECD monitors and periodically revises the CRS, with the most recent revision aimed at collecting information about cryptocurrency holdings. The details of the CRS regime are complex, and its efficacy is imperfect. Nonetheless, it has forced many major financial institutions—including those outside U.S. jurisdiction—to collect beneficial ownership information about all their nonresident customers.
The United States has refused to join the CRS, however, in large part because Washington gets the information it needs under the Foreign Account Tax Compliance Act (FATCA), which instructs U.S. withholding agents to withhold 30 percent from all payments to financial institutions, wherever domiciled, that fail to report their American account holders to the IRS. A 2019 report to Congress from the Government Accountability Office concluded that joining the CRS “would result in no additional benefit to [the] IRS in terms of obtaining information on U.S. accounts.” Although this may be true, participating in the CRS would help the IRS identify beneficial owners who are not American—including those who are Chinese. Joining the CRS to cooperate with trading partners in fighting tax evasion would be a good idea in general. Doing so specifically to counter China’s foreseeable efforts to circumvent future tax reforms is a no-brainer.
There is one remaining loophole that Congress must close. That loophole, which some wealthy American individuals have used to circumvent FATCA and which, in 2020, led to the largest tax-evasion prosecution in U.S. history, is the creation of “captive banks,” or financial institutions formed and controlled for the purpose of evading reporting requirements. Congress’s options in dealing with captive banks are limited. Such banks have no U.S. account holders and so can easily comply with FATCA. And they are organized in jurisdictions that do not participate in the CRS, so joining the CRS would not yield the IRS any additional information about them. The only tool left, therefore, is a blunt one: to require U.S. financial institutions to withhold the maximum 30 percent on any payment to a foreign account not subject to CRS reporting.
This would be a draconian measure. Numerous private companies and trusts set up by wealthy individuals for all sorts of reasons, many perfectly legitimate, would be caught in the dragnet of punitive withholding. Fortunately, there is a way to make such withholding much less draconian—and for most legitimate private investors, at least those not based in China, innocuous. The key lies in the vast size of China’s U.S. investments. Shifting China’s $1.8 trillion in holdings of U.S. portfolio securities into a few captive banks would lead to highly visible new payment streams from the United States. Even if that $1.8 trillion was split into 100 captive banks, each bank would have, on average, $18 billion in assets. A very modest three percent cash-flow rate of return would yield each of them $540 million in annual payments. Creating 1,000 banks would drop this number to a still substantial $54 million. If Congress were to mandate 30 percent withholding on payments to CRS-noncompliant financial institutions when such payments exceeded, say, $10 million a year, it would require CIC and other major Chinese investors to undertake an astounding amount of administrative work to keep any U.S. tax benefits. Congress and the IRS could thus make circumvention sufficiently costly such that China’s only logical response would be to accumulate fewer U.S. dollars—and therefore fewer U.S. assets. This response would in turn force China to reduce the trade surpluses that generated them.
AMERICAN ASSETS IN AMERICAN HANDS
It is difficult to predict with precision what effect eliminating the U.S. tax subsidy would have on Chinese portfolio investment—and therefore on the U.S. trade deficit with China. Some Chinese entities would no doubt accept a lower after-tax return on Treasury securities if the alternative meant not accumulating dollars. But more than a few would look to other markets, whether at home or abroad.
Using an analysis of the effect of dividend-tax policy on foreign portfolio investment produced by the economists Dan Amiram and Mary Margaret Frank, it is possible to estimate the effect of different rates of withholding tax on the stock of Chinese portfolio investment in the United States and the annual bilateral trade balance. Raising the withholding rate from zero to 30 percent would cause China to reduce portfolio investment in the United States by 27 percent—or $515 billion. Assuming a 5.3 percent rate of return (the August 2024 three-month Treasury yield) on that $515 billion, China would, annually, earn $27 billion less in investment income. This would reduce China’s goods-trade surplus with the United States by an equivalent amount, $27 billion—or nine percent of the present surplus level. The relationship between withholding taxes and portfolio investment uncovered by the economists Mihir Desai and Dhammika Dharmapala suggests an even greater tax effect. A 30 percent increase in the withholding rate would lead to an $876 billion reduction in China’s U.S. portfolio investment and a 16 percent decrease in its goods-trade surplus with the United States.
Taxing foreign portfolio investments is not a silver bullet for reducing large and persistent trade imbalances. Raising such taxes will cut imbalances only moderately. Yet it will do so by reducing the impetus for Chinese dumping, without the drawbacks of tariffs. The Trump-Biden tariffs, which were accompanied by a huge rise in the trade deficit, have raised prices for U.S. consumers, damaged the competitiveness of U.S. exporters needing imported intermediate goods, and distorted the sectoral allocation of production in ways that waste scarce resources at home and abroad. Ending tax subsidies for U.S.-bound Chinese portfolio investment, while mitigating the resulting downsides, will raise foreign revenue, support U.S. production, and help keep American assets in American hands.