There’s a particular kind of headache that comes from reading three unrelated news stories in the same week and realising they’re not unrelated at all, and this has been one of those weeks. In truth, this article exists because I couldn’t quite let those stories go, and I’ve broken my own rule of not writing at the weekend because of it. The articles had been rattling around in my head for while, keeping me awake last night, and before I realised what I was doing, this morning, I was deep into research papers, budget tables, news archives and LinkedIn posts, trying to work out whether the numbers being quoted actually added up. The deeper I dug, the stranger the picture became. What follows is the result of several hours of increasingly obsessive fact-checking, prompted in equal measure by frustration at the money governments somehow always find for wars and by the determination of certain tech billionaires to spend vast fortunes attempting to turn themselves into the first Bond villain.

Let’s start with the sky itself, because that’s where this began, with 1.7 million satellites versus the whole of astronomy. The European Southern Observatory has just done something nobody had bothered to do properly before: modelled what happens to astronomy if everything currently proposed for orbit actually gets launched.

The number ESO astronomer Olivier Hainaut arrived at for “acceptable” damage, the point at which satellite interference stays roughly comparable to routine equipment failures is 100,000 faint, naked-eye-invisible satellites.

The number currently on the table, across every proposed constellation from every operator, is over 1.7 million.

That’s not a rounding error. That’s a different universe of consequence, and the specifics are grim: ESO’s Very Large Telescope could lose up to 28% of its field of view to Starlink alone. The Vera Rubin Observatory, arguably the most ambitious survey telescope ever built, stands to lose hours of observing time every single night. And then there’s Reflect Orbital, a company literally planning to put mirrors in orbit to beam sunlight onto Earth after dark, with up to 50,000 of them proposed by 2035. Hainaut’s modelling suggests they could make the night sky itself four times brighter, even from a light-polluted city like Munich.

We are, with a straight face, discussing whether to switch the night sky off. Not metaphorically, actually switch it off!

Here’s where it stops being an abstract problem for astronomers in Chile and starts being a very local one. Meanwhile at home, the same week this ESO analysis landed, the Royal Astronomical Society reported that the UK’s Science and Technology Facilities Council is withdrawing all funding for e-MERLIN, the network of radio telescopes that includes the Lovell Telescope at Jodrell Bank. Jodrell Bank. The place that tracked Sputnik. The place that’s been quietly doing world-class radio astronomy since before most of us were born. It’s now at genuine risk of closure.

That’s not the whole of it, either. The RAS reported the UK is withdrawing from the James Clerk Maxwell Telescope entirely, cutting 20% from the Square Kilometre Array Regional Centre, cutting 20% from the UK’s own contribution to Vera Rubin, yes, the same observatory the satellites threaten to steal hours from, and cutting 40% from the BISON solar monitoring network.

So: the instrument is being switched off from above, and the people who’d use it are being defunded from within, in the same month. If you set out deliberately to sabotage a field of science, it’s hard to imagine a more efficient one-two.

Now for the number that actually made me sit down and do the maths, and look at what we are proposing to spend it on instead, because I’d heard a vague comparison somewhere and wanted to check whether it held up. It doesn’t hold up quite the way I’d remembered. It’s a lot more startling than that.

The US-Israel war on Iran that began in February 2026, Operation Epic Fury, to give it its official name, has, according to the Pentagon’s own figures given to the Senate on 21 July 2026, cost $37.5 billion so far. That’s up from $29 billion in May and $25 billion in April; the number has been climbing by roughly a billion dollars a week. Defence Secretary Pete Hegseth is now asking Congress for up to $70 billion more in emergency funding, as part of an $87.6 billion supplemental request. If Congress approves that request, the total cost attributable to the conflict could exceed $100 billion.

Here’s the part that stopped me. NASA doesn’t actually cost all that much by federal-government standards. Its annual budget has been remarkably steady at around $24-25 billion a year. The $37.5bn already spent on the Iran war is therefore equivalent to roughly eighteen months of NASA: every mission, every telescope, every ongoing operation, every payroll. Already spent, already gone, no projections involved.

Add the funding now being requested and the comparison grows to roughly four years and four months of the entire US civilian space programme. Say it slowly: four-and-a-bit years of the entire United States space programme, potentially gone in five months of a war that hasn’t produced an observatory, a scientific instrument, or a single piece of usable knowledge about the universe we live in. But hey, that’s OK, it has put the cost of living through the roof.

The arithmetic is straightforward. Add together 2026 and roughly half of 2025 and you reach the equivalent of the $37.5bn already spent on the war. Continue through 2024, 2023 and part of 2022 and the total reaches roughly $107.5bn. Same source data, same calculation, just expressed in time rather than dollars.

I’m not usually one to get on my soapbox, and I do try not to be the “billionaires bad guy,” mostly because it’s boring and it’s usually not that specific a complaint. But specificity is exactly what we have here. We’re not talking about an abstract injustice, we’re talking about a precise, sourced, comparable set of numbers. A dark sky and a functioning radio telescope at Jodrell Bank, for the price of what a war spends in a matter of weeks. Actual, unglamorous, patient scientific research, the kind that doesn’t make anyone rich, doesn’t come with a rocket launch livestream, and doesn’t get you invited on a podcast, going begging for a fraction of a fraction of what gets found overnight for other purposes.

If there’s a genuine, defensible case for where the enormous fortunes of the “let’s conquer space” tech-billionaire set should go, it’s here: into keeping Jodrell Bank’s lights on, not into another few hundred thousand satellites and orbital mirrors marketed as innovation. And if there’s a case for where a nation’s money should go, the maths above makes it for me. NASA gave us the Moon landings, Hubble, and forty years of thinking a bit harder about our place in the universe, and it did all of that on this budget. The full Iran war bill, once the current request lands, could run past four years and four months of that budget. But you don’t even need the bigger number to make the point, eighteen months of NASA, gone in five months of a war that won’t produce a single new fact about a single distant galaxy.

We can afford to look at the stars. We just keep choosing not to.

A quick note on the sources: The comparisons made in this article are drawn from publicly available sources, including ESO’s satellite-impact study, the Royal Astronomical Society’s reporting on UK astronomy funding, Pentagon cost figures reported by Reuters, the BBC and The Guardian, and NASA budget data. The references are included so readers can verify the numbers for themselves and see exactly how the conclusions in this article were reached. Regular readers will know I don’t usually include a source list like this, but because this article grew out of several hours of fact-checking and number-crunching, I thought it was only fair to show my working.

Sources:
ESO, “Beyond the limit”: one million satellites and mirrors in space pose grave threat to the night sky — the original ESO/Olivier Hainaut study
New Atlas, “1.7 million planned satellites to have ‘devastating’ impact on astronomy” — corroborates the 28% VLT field-of-view figure
Royal Astronomical Society, “Mass cuts and potential Jodrell Bank closure ‘devastating’ for astronomy”
Reuters, “War in Iran has cost the US $37.5 billion so far, Pentagon says” (21 July 2026)
Reuters, “US war in Iran has cost $29 billion so far, Pentagon says” (12 May 2026)
The Guardian, “Hegseth tells Senate war on Iran has so far cost $37.5bn” (21 July 2026)
BBC, “Iran war has cost US $37.5bn so far, Hegseth says” (21 July 2026)
Wikipedia, “Budget of NASA” (figures sourced to The Planetary Society’s historical NASA budget dataset)