I Want Free Wednesdays at 40. I Can Buy One This Friday.

Three weeks ago I wrote that the thing at the end of our plan isn't a beach. It's a classroom, a maths timetable and free Wednesdays. Then I did the obvious thing I had somehow never done. I counted the Wednesdays.

There are about 624 of them left. Every one gets the same answer from the plan: not this one. All of them at once, in twelve years, in a single delivery. Meanwhile a four-day week has been sitting there the whole time, a thing I could ask for in a fortnight at a price I could compute in an afternoon with the projection tool I built for myself.

I never priced it. Not once. That is either discipline or avoidance, and I couldn't tell you which without the number, so last week I finally ran it.

It came out cheaper than I expected, and it contained one result I checked three times because I assumed I'd broken the model.

The plan sells one product and delivers it all on one day

Take the spreadsheets off financial independence and it is selling exactly one thing: hours you get to decide about. The money is a receipt. Nobody wants 25× annual spending, they want the Tuesday afternoon that 25× buys.

What's strange is the delivery schedule. You pay in for twelve years, receive nothing for twelve years, and then on one ordinary day in the 2030s the entire order arrives, permanently, all at once. I can't think of another purchase in my life that works like that. We don't buy holidays this way.

A four-day week is the same product on instalments. Same thing in the box, sold one Wednesday at a time, delivered next month. The FIRE world already has names for partial versions of this, and I've written up barista and coast FIRE before, but both of those are about changing the job or stopping the saving. This one is about staying exactly where you are and buying back a day. For years my plan has declined that offer without ever asking the price.

Refusing a thing you never priced isn't discipline. It's just a habit wearing discipline's jacket.

There are two four-day weeks in 2026, and only one of them has a price

This is where most writing on the subject quietly cheats, so let me split them before I use a statistic.

The loud one is the employer-run pilot. The trial published in Nature Human Behaviour in July 2025 covers 141 organisations and 2,896 employees across six countries and finds real improvements in burnout, job satisfaction and self-rated health against controls. The UK pilot run by Autonomy put 61 companies and roughly 2,900 workers through six months: revenue up 1.4%, staff turnover down 57%, 92% of the companies carried on afterwards.

Every one of those numbers held pay at 100%.

Which means none of them are evidence about the decision you and I actually face. If your employer is running a 100-80-100 pilot you aren't making a trade, you're receiving a gift, and you should take it before somebody in finance reads the P&L differently. The individual version is a completely different transaction: you walk into a one-to-one and ask for 80% of the week at 80% of the pay. There is no trial for that. There is no dataset of people who did it deliberately as a financial independence decision. Anybody quoting pilot productivity data at you to justify your own pay cut is mixing up two things that share a name and nothing else.

And a chunk of this is already ordinary somewhere. An American works around 1,800 hours a year. A German works 1,332.

Data: OECD, average annual hours actually worked per worker (2025), via compiled table. Includes part-time workers.

That gap is 468 hours a year, twelve extra forty-hour weeks. A good slice of what an American reader is saving twelve years to buy, a German already has and doesn't think about. Same arithmetic on both sides of the Atlantic, very different social price for asking.

What one free Wednesday costs, in years

Here is the part people's instinct gets wrong, and my instinct got it wrong too.

A 20% pay cut is not a 20% hit to your plan. It's much worse, because your spending doesn't take the cut with you. Say you earn 100 and spend 55, so you save 45. Drop income to 80 and the spending stays at 55, so now you save 25. A fifth off your income took nearly half of your engine. The currency doesn't matter. The ratios do.

So I modelled it for a saver further into the run than we are: portfolio already at ten times annual spending, 5% real return, financial independence at 25× spending, income cut exactly 20%, spending completely flat.

Modelled by me, not measured data: 5% real return, FI = 25× annual spending, starting assets = 10× spending, income cut 20% at an unchanged hourly rate, spending held flat, year-end contributions, no tax.

At a 45% savings rate it's 32 months. At 60% it's 21. At 25%, the worst case on the chart, it's 49.

Under two years at the top end, just over four at the bottom. That's the price of a permanent free Wednesday for the rest of your working life, for someone already a decent way into the run. I would have guessed a decade. I'd have said a decade out loud, confidently, standing in the hardware store.

That number is also the entire reason I think the long middle stretch is where this decision belongs. The boring middle of FIRE is a decade of doing the identical thing while the portfolio quietly takes over the work of compounding. It's the cheapest place on the whole curve to buy a day, and it's the exact stretch where everybody grits their teeth instead.

The savings rate decides this, and most advice gets it backwards

Read what's written about this and you get the same framing everywhere: can you afford the pay cut? Asked as a question about how big your salary is. Somebody on Bogleheads asks whether to take a $60,000 cut for a four-day week and the replies pile in, nearly all of them reacting to the size of that number.

That's the wrong variable. Two people on wildly different salaries with the same savings rate get exactly the same answer from this model, because it's ratios all the way down. What decides it is the gap between what you earn and what you spend, which is the same lever that does more work in your plan than your investment returns ever will.

Then the result that made me rerun the whole thing.

Modelled by me, not measured data: same assumptions as the chart above, but starting invested assets = 0.

Cut your income by a fifth and cut your spending by a fifth, and your financial independence date does not move. Not "barely moves". Does not move, to the decimal, at every savings rate I tested.

I checked it three times. It's correct, and it's obvious in hindsight: every quantity in the model is a ratio. Your savings rate is unchanged, your target is a multiple of spending so it shrinks by the same fifth, and nothing is left that knows how big the numbers used to be. Shrink both sides of the ledger together and the date is blind to it.

Which reframes the whole decision. The pay cut is only expensive if your spending refuses to move. The Wednesday is not priced in euros, it's priced in how discretionary your life actually is at the margin.

Now the honest correction, because I'd be doing exactly what I just accused everyone else of doing. Almost nobody delivers that. A free day genuinely removes a commute, some childcare, some convenience food and a chunk of the compensatory spending you do because you're knackered. It also adds hobby costs and daytime costs, and most people who assume minus twenty percent deliver minus five. The middle line on that chart, spending down 10%, is the realistic one, and it's still a big improvement on doing nothing.

And then the sentence I like least in this article. On that same zero-start model, below roughly a 22% savings rate with spending held flat, the pay cut swallows the entire surplus and financial independence never arrives at all. At 25% it turns a 32-year path into a 61-year one. The reader who most wants a Wednesday off is, almost by definition, the one who cannot buy one. Every cheerful "just go to four days, life's short" post is aimed squarely at the people it would hurt most.

The ledger where I lose this argument

I'm not going to hide this one in the last section, because it's the best argument against everything above and I only found it by building the model properly.

Run the full course for that mid-journey saver at a 45% savings rate. Going to four days buys 618 free Wednesdays before the date, which is almost exactly the number I opened this article with, and that isn't a coincidence, it's the same dozen years counted the same way. But it also pushes the date out by 32 months, and every one of those months was going to be entirely free. Net across the whole run, they work about 72 days more than if they had never touched their schedule.

So the claim is not "you get more free time." On a lifetime ledger you get less. The claim is "you get free time now, and you pay for it with free time later." That's a preference, not a free lunch, and I think it's a defensible one. A Wednesday at 28, with a small child in the house and knees that still work, is not the same good as a Wednesday at 58. Compounding runs on health and on your kids' ages too, and both of those curves run the other way from your portfolio. If you want to argue that 618 Wednesdays in your thirties beat 72 days back in your fifties, I'll help you argue it.

What I won't accept is anyone pretending the trade is free. It isn't. It's a currency exchange between two versions of yourself, and one of them isn't in the room.

Data: Eurostat, part-time and full-time employment statistics, 2024. The comparable US figure is 17.5% (BLS Table A-9, August 2026), but Eurostat uses self-defined part-time while the BLS uses under 35 hours — comparable in spirit, not in method.

That chart also kills the idea that any of this is exotic. 38.6% of Dutch workers are part-time, 28.9% of Germans, and across the EU 31.7% of employed women with children, rising to two thirds or more in Austria, the Netherlands and Germany. The trade I've spent a week modelling as an extreme financial manoeuvre is what a very large number of European parents simply do, without a spreadsheet, and have done for decades.

What the delay might genuinely be buying

Now the other side, properly, because there are real reasons to keep working five days and I don't want to build a version I can knock over.

Start with the assumption I made. My model trades cleanly pro rata: 80% of the time, 80% of the pay, same hourly rate. The Economic Policy Institute puts the US part-time hourly wage penalty at 19.8% even after controlling for demographics, education, occupation and industry, plus another 5.5 percentage points on non-wage benefits. If your rate takes a hit as well as your hours, an 80% schedule lands closer to 65% of your income and every number above gets worse. I gave the trade the benefit of the doubt.

Then there's the slope, which matters more than the level. A 20% cut priced once is arithmetic. A 20% cut that also takes you off the track where your income was growing in real terms is a flatter curve for twenty years, and no chart is going to show you that. Skills and career are the engine that makes a high savings rate possible at all. This is the objection that actually stops me.

And the option runs one way. Going to 80% is a conversation. Going back to 100% at the same seniority, after two years of somebody else getting the interesting projects, may not be on the menu. You aren't buying a Wednesday. You're buying the right to keep buying Wednesdays, and possibly selling the right to stop.

One objection isn't financial at all. Sharif, Mogilner and Hershfield found wellbeing rises with discretionary time up to about two hours a day, plateaus, then falls off past five unless the time goes into something active rather than a screen. The first free Wednesday is a holiday. The twentieth is a Tuesday with worse television. Same rule I landed on when I finally put a line in our budget for buying back an hour: a bought hour needs a named destination and a review date, or it's just convenience with a price tag. A bought Wednesday is that at fifty times the scale.

One category I'll name and then leave alone, because it's different everywhere. Most European pension and unemployment entitlements are earnings-linked, so a permanent income cut trims the state floor under your private plan. In the US, health insurance and workplace retirement access turn on hours thresholds that can make the fourth day cost far more than a fifth of your pay. Different plumbing, same direction: check yours before you believe my chart.

Why I still work five days

Three reasons I'd defend, and one I wouldn't.

The boring one first. We're a one-income household right now, my wife is home with our daughter, and voluntarily removing a fifth of the only salary in the building this year would be a special kind of daft. That one isn't a philosophy, it's a date on a calendar.

Second, I like the job. Not in the way people say it in interviews. I'm not counting down to anything, which is the whole reason our plan reads the way it does rather than as an escape route, and hating your job is not a plan. If you're miserable at work, a four-day week fixes a fifth of a problem you should be fixing completely. Third, I'm on the steep part of the income curve, and the slope argument above is the one I believe most.

And the one I wouldn't defend. For years the answer was no, and I never once quoted myself a price. I let a default masquerade as a decision, in the one area of our finances where I model absolutely everything else. That's not discipline. That's the same avoidance I write about in other people.

So here's what would flip it, written down where I can be held to it. When we're a two-income household again. When the income curve flattens, which I'll notice because I track it. When the portfolio crosses roughly ten times our annual spending, which is where my own chart says a Wednesday gets cheap. And when I can say what the day is actually for, which for me is not a lie-in. I've claimed in public that I want to teach maths at 40 and that the claim is a hypothesis I owe myself a test of. A term of guest lessons fits neatly into a Wednesday. If I'm still refusing to buy one when all four of those are true, I'm not saving for freedom, I'm collecting it.

Run the four-day week number before you defer it

This takes an evening, not a spreadsheet habit.

Work out your real savings rate over twelve months, not last month. Then the post-cut one, on the new income: 0.8 times income minus spending, divided by 0.8 times income. A 40% saver becomes a 25% saver. A 20% saver becomes a zero saver. Be honest about what your spending does, and if you think it drops a fifth, go and name the categories out loud. Then run both versions to the date and read the gap in months. Convert that into fully free days and hold it next to the Wednesdays you'd be buying. It's an exchange rate, and you're allowed to take a bad one on purpose as long as you saw it.

The projection in the tracker I build is the only place I've found where the two versions of your life sit side by side and you can watch one of them slide while you drag the other. That's most of why it exists.

I'm still going to be at my desk on Wednesday. But a four-day week has a price tag on it now instead of costing nothing, and refusing something at a quoted price is a different act from refusing it out of habit and calling the habit a virtue.

Our long money date is in January. This is going on the agenda with a figure next to it, and I'll have to say out loud what I think a run of Wednesdays while she's small is worth against a pile of fully free days when she's grown. On the model that trade is 618 against 72. I already suspect I know the answer. I've just never had to defend it with a number before.