Garvan Walshe is a former National and International Security Policy Adviser to the Conservative Party. He runs Brexit Analytics.

What connects Donald Trump, the Kurdish retreat from Iraq’s oil-rich Kirkuk region, a Syrian government push against some remnants of ISIS, a bloody, expensive war by the Saudis in Yemen and the continuing strength of Hezbollah in Lebanese politics?

The answer, of course (the reference for Hezbollah might have been a clue), is Iran. Trump, as was his way in business, never regards an agreement as final. There he gained a reputation for never paying suppliers in full, and seeing how much they were willing to drag out with litigation. He behaves the just same way with friends (demanding NATO members pay more on defence if they want American support) as with foes.

For months a battle has been raging within his administration over the Joint Comprehensive Plan Of Action (JCPOA) to control Iran’s nuclear ambitions. Trump wants to “decertify” Iranian compliance, and begin the process that would lead to the US pulling out of the agreement. The administration’s “adults” – Rex Tillerson, Secretary of State, and James Mattis, Secretary of Defence – are opposed. The compromise reached was for Trump to ask Congress to do this for him.

By doing so, he has damaged reputation of the United States, split Washington from its allies, and strengthened, not weakened, Tehran, which will chalk up another miscalculation of which it can take advantage. That much is done. What matters is whether Congress will now take out the rubbish for him.

The JCPOA works by tightening loopholes in the international nuclear non-proliferation regime that have allowed Iran to get very close to developing a bomb without being caught breaching the terms of the Non-Proliferation Treaty. The new measures include forcing Iran to eliminate its stock of medium-enriched uranium (which can quickly be converted into the high-enriched stuff needed for weapons) and cut the number of enrichment centrifuges it can operate. Iran’s justification for continuing with this is the transparent fiction that it is engaged in civil nuclear fission research, framed as a matter of national pride. This obviously makes no sense for a country with huge oil reserves and abundant sun, mountain rivers and wind. Plentiful energy can be obtained right now from fossil fuels while pride in Iranian science would be better served by research into exciting renewable energy instead of expensive dangerous and ever-so-20th-century fission.

Nevertheless, the effect of the agreement is to delay Iran’s acquisition of nuclear weapons. The relevant question is not whether it will stop Iran’s nuclear programme for good – it won’t – but whether it is better than alternative, military means of non-proliferation.

To see whether it is, Congress should apply a test I first heard from Binyamin Netanyahu – not, shall we say, a known friend of Iran. This is not the cartoon drawing of a bomb with a lit fuse which he presented at the UN General Assembly, but the image of Iran’s two clocks that he used to use the last time Iran had, as it does now, a reformist president.

Iran, he used to say, ran on two timelines: a nuclear clock, how long would it be before its nuclear research effort bore fruit? And a political clock: how long could the regime maintain control of an unhappy people long tired of repressive, inward-looking theocracy, and which was nearly overthrown by them in 2009?

Open military intervention, with all its attendant risks, is thought to be able to set the programme back at most about five years. The JCPOA, if it is upheld, by around fifteen. And war, like it did in the 1980s, when Iraq attacked Iran, will in all probability strengthen the hardliners of the Revolutionary Guards, who answer directly to Iran’s executive Supreme Leader, and weaken its elected institutions.

This will be of particular importance in the next few years, because Ali Khamene’i, the current Supreme Leader, is ailing, and proposals are being aired to replace him with a council, or even leave the post vacant altogether. Political change in Iran is coming, and it would be a major strategic mistake to snuff it out.

The Netanyahu Test recommends sticking with the agreement. It slows down the nuclear programme, giving time for political change to occur. Its collapse would allow Iran to resume its programme impeded only by sanctions that have been shown to be of limited effectiveness, and increase the chances of a counter-productive war breaking out.