Even though the administration was right to deny the waiver, waivers are a vital part of the sanctions process. They help governments target sanctions effectively, giving them flexibility to respond to changing circumstances on the ground. There are at least two reasons that this ability to target and respond flexibly is important. The first is strictly moral: It is wrong, for example, to harm Iraqi children caught in a dispute between the U.S. and Iraqi governments. The second is more pragmatic: Sanctions effectively aimed at decision makers puts pressure on them, reduces their ability to ally with fellow targets, and reduces opposition to the sanctions on humanitarian grounds. Waivers help to ensure that comprehensive sanctions are more targeted and shaped. Democracies—including limited democracies like apartheid South Africa—are more responsive to sanctions than are dictatorships, and pressure on voters becomes pressure on politicians.

These are lessons hard learned, particularly following Iraq’s 1990 invasion of Kuwait. Initially, the embargo imposed by the UN Security Council on Iraq was comprehensive, rather than targeted, with devastating effects on Iraq’s economy. It wasn’t until 1995 that the U.S. and Iraq came to an “oil for food” agreement to relieve innocent civilians. The agreement allowed Iraq limited oil sales to fund humanitarian imports, UN weapons inspections, and reparations for victims of its 1990 invasion. The Security Council’s initial optimism that comprehensive sanctions could be an inexpensive way of conducting foreign policy, gave way, instead, to abandoning comprehensive sanctions in favor of “smart” or targeted sanctions.

Thus, with Iran in the 2000s and 2010s, the United States and its European partners sought to learn lessons from Iraq. During the toughest period of sanctions imposition, the United States expanded the definition of medical goods permitted to go to Iran without a license in order to prevent undue suffering on the part of the Iranian population. This waiver, which required no quid pro quo from Iran, sought to alleviate Iranian civilian suffering of the Iranian people notwithstanding the underlying dispute, while also weakening the perception that Iranians were suffering due to sanctions.

Waivers can also be designed to advance very specific policy objectives. For example, in 2013, the United States allowed the export of personal telecommunications devices to Iran to help Iranian civil society groups communicate with one another. In January 2017, the United States waived the sanctions against Russia’s intelligence service, the FSB, to allow U.S. companies to import personal IT goods into Russia. The FSB was responsible for granting Russians import licenses, so the waiver was necessary to avoid complicating innocent business between the United States and Russia.

In the UN, sanctions waivers are written into resolutions or granted by the Security Council, so their application is relatively straightforward, if politically complicated. For the United States, there are three general ways to waive sanctions. The first is via statute, whereby Congress either identifies permitted conduct or authorizes the president to make case-by-case decisions according to rules that Congress sets.

The second is via general licensing, whereby the Treasury identifies and defines exempt conduct (such as with humanitarian goods to Iran) on its own initiative—no application is necessary to begin this process. Any U.S. person can take advantage of such general licenses (and, by extension, foreign people can generally assume that U.S. government support exists for the business implicated).

The third type of waiver is the one that ExxonMobil sought: a specific license for an identified transaction that sanctions would otherwise prohibit. While the terms of what can be applied for are not pre-specified, allowing a company to apply for any exemption they want, they should not expect to secure such a license if it is directly at odds with U.S. policy.

There is no particular pattern for licenses being granted or denied, and there is limited data available due to privacy constraints on the Treasury Department. Some information is available via the Treasury’s biannual and quarterly reports under the 2000 Trade Sanctions Reform Act (TSRA), which focused on licensing agricultural and medical goods to Sudan and Iran. A cursory review of these reports shows that, for most of the past 17 years and for these types of goods, the Treasury has licensed approximately two thirds of the applications, and issued only a handful of denials.

ExxonMobil’s application is argued on business grounds rather than humanitarian or policy grounds—not the usual criteria for successful applications. ExxonMobil’s implicit view that the sanctions will not bring about Russian compliance with U.S. policy goals if European companies can flout them has undeniable appeal: maintaining a broad, international coalition requires real commitment and diplomatic skill, and—even then—may fail. Why should U.S. companies lose out on profitable joint ventures to support European security that even European companies seem willing to jeopardize?

This presents an opportunity for the Trump administration to go beyond merely denying the application. Sanctions become ineffective as policy tools when left to drift, and these sanctions have clearly entered those waters. The Trump Administration should seek to revisit and re-energize the international response to Russia’s aggression. Doing so would be a very welcome sign of statecraft from an administration that has allowed very serious doubts to linger about its commitment to international diplomacy and long-standing U.S. allies.

Richard Nephew is a senior research scholar at the Center on Global Energy Policy at Columbia University. Colin Rowat is an economist at the University of Birmingham.