Collective Punishment With Better PR

The word that lets us sleep

Some words are chosen because they describe a thing. Others are chosen precisely because they don’t. “Sanctions” sits firmly in the second drawer, alongside “collateral damage”, “enhanced interrogation” and that old favourite of every corporate restructure I ever had the misfortune to sit through, “efficiencies”. It’s a word with its tie done up and its shoes polished. It sounds like something a magistrate hands down after a stern look over half-moon glasses, not something that empties a hospital pharmacy in a country most of us couldn’t find on a map without a fair bit of squinting.

I’ve been chewing on it for weeks now, which is one of the hazards of living in a cave on a Spanish hillside with a great deal of quiet and a news diet that arrives in fragments. Fragments have a habit of sticking. Another round of measures here, a fresh package there, a minister at a lectern explaining that this time the pressure will be targeted, proportionate and firm. Nobody ever quite says who it lands on, and the question I keep arriving at, usually somewhere around the second coffee, refuses to go away.

If a policy knowingly causes hunger, illness and poverty among ordinary people in order to change the behaviour of the people who govern them, what exactly do we think we’re doing? Because from where I’m sitting, it looks an awful lot like collective punishment with better PR.

The rule they taught me first

I spent a good chunk of my working life in the RAF Police and then in security management, and if those years drilled one principle into me, it’s this: you deal with the person who did the thing. Not his neighbours, not his children, not the poor sod who happens to run the corner shop on his street. Individual responsibility isn’t some soft liberal nicety, it’s the load-bearing wall of every justice system worth the name. The moment you start punishing people for what somebody else did, you’ve stopped policing and started doing something else entirely.

The irony is that the military understands collective punishment better than most. Anyone who’s been through basic training knows the old routine… one recruit’s bed isn’t squared away, so the whole flight goes round the block again. It works, after a fashion. It also breeds resentment, turns people against each other and teaches nobody anything about making a bed. It’s tolerated on a parade square because, crucially, it stops at the edge of the parade square.

Out in the real world, the grown-ups wrote the rule down. Article 33 of the Fourth Geneva Convention is short enough to fit on a beer mat: nobody protected by it may be punished for an offence they didn’t personally commit, and collective penalties are prohibited. The drafters had the reprisal killings of two world wars in mind, whole villages made to pay for one ambush. Anyone who has sat through a Law of Armed Conflict lecture, eyes glazing over in a stuffy briefing room, has heard some version of it.

Now, I’ll be scrupulously fair here, because a good lawyer would be. Article 33 governs civilians in the hands of a party to a conflict or an occupying power, and economic sanctions imposed from thousands of miles away don’t sit neatly inside its wording. The people who design them know that perfectly well. But I’m not much interested in whether sanctions slip through a legal gap. I’m interested in why the principle exists at all, and whether we’ve simply found a way to do the forbidden thing at a distance, with a spreadsheet instead of a firing squad.

The comfortable story

The official version goes something like this. A regime does something monstrous, and the world can’t simply shrug and carry on. So rather than sending in the troops, we squeeze: freeze the assets, restrict the trade, cut off the finance, and make life miserable for the people at the top until they change course. No body bags, no flag-draped coffins on the evening news, no awkward public inquiry a decade later. Clean.

I don’t doubt that, for some of the people who design these things, that really is the intention. But I spent far too many years writing incident reports to confuse intentions with outcomes. What matters, in the end, is what happened, to whom, and who knew it would.

So who actually pays? It’s rarely the people in the presidential palace, the generals’ compound or the penthouse in a friendly capital with a refreshingly relaxed attitude to where money comes from. Those people have stockpiles, smuggling networks, second passports and a cousin in import-export. Sanctions are an irritation to them. To the nurse who can’t get insulin, the father watching the price of bread double in a month, or the child whose cancer ward has run out of the drugs that might have kept her alive, they are something else altogether.

There’s a nasty little twist in the tail, too. Squeeze an economy hard enough, and you hand the regime control over whatever’s left. Who decides who gets the rationed flour? Who runs the black market in medicines? Not the dissidents, I promise you. Sanctions have a way of making ordinary people more dependent on the very government they were supposed to loosen its grip on, while handing that government a ready-made foreign villain for every empty shelf.

“We think the price is worth it”

If you want to know where this road leads, you don’t need a theory; you need 1996. After Iraq invaded Kuwait in 1990, the UN imposed some of the most comprehensive sanctions ever devised. Iraq wasn’t permitted to buy materials to rebuild its water supply, farming or once-respected medical system until 1998, the best part of eight years in. Shipments of syringes were held up for half a year over fears they could be turned to anthrax production. Plastic bags for blood transfusions were restricted. That’s what “pressure” looks like when you zoom in.

In May 1996, Lesley Stahl of 60 Minutes put a figure to Madeleine Albright, then America’s ambassador to the UN: half a million children dead, more than at Hiroshima. Was the price worth it? Albright called it a very hard choice, and then said, in so many words, that they thought it was. She didn’t dispute the number. She weighed it, pointed to the concessions Saddam had made on weapons and on recognising Kuwait, and found the trade acceptable. It’s a moment that ought to be taught at every staff college in the world, not because one diplomat fumbled a television interview, but because she said out loud what the policy had been quietly assuming all along.

Fairness matters more than ever when the subject is this ugly, so here it is. The half-million figure was a rough rendering of an early Food and Agriculture Organisation estimate, and later work by Richard Garfield at Columbia put excess deaths among under-fives considerably lower, somewhere between roughly 106,000 and 227,000. Albright herself later wrote that her answer had been a mistake she regretted the moment she’d said it. But sit with the lower figure for a minute. A hundred thousand dead children, at the conservative end, and still the policy rolled on.

Meanwhile, the man it was all aimed at stayed in his palaces, ran the ration system, controlled the smuggling and used the suffering as propaganda. It took a full-scale invasion to remove him in 2003. The sanctions never did the job they were sold to do. They just did the other job, the one nobody put in the brochure.

Smart sanctions and the frightened bank

The modern answer to all of this is that we’ve learned our lesson. These days we have smart sanctions. Target the elite, freeze the oligarch’s yacht, block the general’s bank account and leave the shopkeeper alone. It’s a lovely idea, and it is genuinely better than the blanket embargo that flattened Iraq. The trouble is that an economy isn’t a chessboard. You can’t lift one piece off without the whole board shifting.

What actually happens is fear. I spent years in senior security management, and I know exactly how risk-averse institutions behave when the penalty for getting it wrong is enormous, and the reward for getting it right is precisely nothing. They don’t calibrate. They run. A compliance officer looking at a modest payment to a hospital supplier in a sanctioned country isn’t thinking about the patient at the other end. He’s thinking about the fine, the headlines and his pension, so he says no. The shipping firm says no. The insurer says no. Nobody broke a single rule, and everybody followed the incentive.

There’s a name for this. Alena Douhan, the UN’s Special Rapporteur on unilateral coercive measures, calls it over-compliance: banks and businesses going well beyond what the law requires and cutting clients off to avoid risk rather than manage it. After visiting Venezuela in 2021 she described humanitarian exemptions as “lengthy, costly, ineffective and inefficient”, against a backdrop of shortages in electricity, water, fuel, food and medicine. She’s raised the same alarm over Iran and Syria. Lawyers who work in this area point out that humanitarian payments are sometimes refused simply because they’re too small to be worth a bank’s compliance costs. A child’s medicine, it turns out, isn’t worth the paperwork.

To give credit where it’s due, the UN Security Council did sort out its own patch. In December 2022 it passed Resolution 2664, a standing humanitarian carve-out from asset freezes across every UN sanctions regime, replacing the old patchwork of case-by-case exemptions. That was a real step forward, and the people who fought for it deserve the credit. But it covers UN sanctions. Much of the squeeze these days comes from unilateral measures imposed by individual states or blocs, chiefly the United States and the European Union, and an exemption on paper still has to survive a nervous legal department before it becomes medicine on a shelf.

The numbers, and that word “unintended”

Here’s the figure that stopped me cold. In July 2025, The Lancet Global Health published a study by the economists Francisco Rodríguez, Silvio Rendón and Mark Weisbrot, which tracked age-specific death rates across 152 countries over five decades. Their estimate was that unilateral sanctions, the ones imposed by the US or the EU, were associated with around 564,000 deaths a year between 2010 and 2021. That’s more than five times the average annual number of people killed in battle over the same period, and in the same ballpark as some estimates of war’s entire yearly toll, civilians included. Across the whole 1970 to 2021 period, children under five accounted for 51 per cent of those deaths.

I’m not going to wave that number about like a placard, because it deserves better than that. It’s a statistical estimate of excess mortality, not a body count. It can’t tell you which individual death a sanctions regime caused, and it can’t perfectly untangle sanctions from everything else going wrong in a country living under them. Two of the authors are associated with the Centre for Economic and Policy Research, which has long been critical of sanctions, and critics will point that out. Fine. Halve the number if you like. Quarter it. You’re still left with a figure that, had it come from missiles, would have us marching in the streets.

And this is where the language does its quiet work. When sanctions kill, we call it “unintended consequences”. What a marvellous phrase. It lets everybody off the hook at once. The politician didn’t intend the hunger, the banker didn’t intend the child to miss her chemotherapy, and the official didn’t intend the anaesthetic to run out halfway through an operation. Perhaps none of them did.

But I’ve written enough statements, and read enough of other people’s, to know the difference between an accident and a foreseeable outcome. If you adopt a policy knowing civilians are likely to suffer, and you keep it in place for years after the evidence has come in, “unintended” stops being a moral position and becomes a legal defence. In my old line of work, carrying on regardless after you’ve been warned had a name, and it wasn’t “accident”.

Do they even work?

You might forgive a great deal of this if it worked. If the squeeze reliably brought tyrants to heel, there’d at least be a grim arithmetic to argue over, lives lost now against lives saved later. So does it?

The most quoted research, by Gary Hufbauer, Jeffrey Schott and Kimberly Elliott, looked at sanctions episodes stretching back to the First World War and found they contributed to some measure of success in roughly a third of cases. That’s the optimistic reading. The political scientist Robert Pape went through the same cases and argued that only a handful genuinely deserved the label, putting the real figure below five per cent. When other researchers applied the Hufbauer yardstick to ten sets of targeted sanctions from the 1990s, they scored just two as successes.

Somewhere between one in three and one in twenty, then, depending on who’s marking the homework. Hardly a precision instrument. And notice which way the failures tend to run. Pape’s point was that pressure from outside often strengthens a ruler’s nationalist credentials rather than weakening them. Give a dictator a foreign enemy to blame for every empty shelf, and you’ve written him the best speech he’ll ever give. Cuba has lived under an American embargo for about as long as I’ve been alive and is still run by the same party. North Korea has been sanctioned to the eyeballs for decades and keeps building missiles regardless. The list of regimes that outlasted the pressure is long. The list of ordinary people who didn’t is longer.

There’s a more cynical reading too, and I can’t quite shake it. Perhaps sanctions don’t exist primarily to change the target at all. Perhaps they exist so that governments at home can be seen to be doing something. A press release with teeth, if you like… It’s just that the teeth are always somewhere else.

So what instead?

This is where I have to be honest, because the easy version of this article writes itself and it’s wrong. I don’t think the answer is “never sanction anyone”. Sometimes a regime commits atrocities, and the rest of the world carrying on with business as usual isn’t neutrality; it’s complicity with a receipt. I’ve also spent enough of my life around the military to know that force is rarely the cleaner option. Bombs are quicker, less discriminating and leave a mess that outlives everyone who ordered them.

So the choice was never sanctions or nothing. It’s what kind, aimed at whom, for how long and judged by what. Here are a few questions I’d like every minister to answer before the next package is announced, rather than after:

What, specifically, has to change for these measures to be lifted? If there’s no clear exit, it isn’t pressure; it’s a siege.
Who is measuring the harm to civilians, independently, from day one, with the power to force a review? We count the oligarchs’ frozen yachts readily enough. Who’s counting the empty incubators?
Are food, medicine, water, fuel and humanitarian payment channels protected before the policy begins, with banks given written assurance rather than left to guess and panic?
Does this regime need civilian suffering in order to “work”? If it does, it isn’t a smart sanction at all. It’s the old blunt one in a nicer suit.
Would we be prepared to say the expected human cost out loud, on camera, the way Albright did?

None of those are comfortable questions. Every one of them is better than the question we usually ask, which tends to begin and end with how tough the announcement makes us look.

Violence that looks like paperwork

Here’s what I keep coming back to, sat outside of an evening with the hillside going dark around me. Sanctions let powerful countries feel morally active without ever paying the bill themselves. The cost lands somewhere else, on people with no vote in our elections and no seat at our tables, people who never chose their government and, in plenty of cases, have been risking their necks to oppose it. We get to say we acted. They get to live with the consequences, or not live, as the case may be.

That doesn’t make every sanction indefensible. It ought to make us a great deal less smug about them. Because if our idea of a humane foreign policy is one that avoids bombs while quietly tightening the screw on an entire population until something gives, then perhaps we haven’t rejected violence at all. We’ve simply found a way to file it.

I do not have an answer for this, as such, and I’m wary of anyone who claims to. What I have is the principle I was handed at the start of my career, the one that says you deal with the person who did the thing. Somewhere between Whitehall, Washington and Brussels, that principle seems to have been mislaid. I’d quite like someone to go and look for it, before the next child pays for a decision taken by a man she’ll never meet.