I Paid for Everything in Cash for Two Weeks and It Was Not the Lesson I Expected

The cash experiment is a standard piece of advice and the stated mechanism is emotional: physically handing over notes hurts in a way that tapping a card does not, so you spend less. I did two weeks of it, spent about eighteen percent less than my usual fortnight, and I am fairly confident the emotional mechanism had almost nothing to do with it.

What actually happened was that I had to plan. Carrying cash means deciding in the morning how much to take, and deciding how much to take means thinking about what the day contains. I was not spending less because paying hurt; I was spending less because I had already thought about it before I left the house, and the thinking was the intervention.

The evidence for this, as far as a sample of one goes, is where the reduction landed. It was not spread evenly across purchases as an emotional friction effect would predict. It was concentrated almost entirely in unplanned stops — the coffee I had not intended, the shop I went into because I was passing. Planned purchases were unchanged. On the two days where I took out a generous amount and had not planned, I spent normally.

This matters because the cash version has real costs. It is inconvenient, some places do not take it, and it makes tracking harder rather than easier — I lost a week of accurate records because I could not remember what I had spent £12 on. If the working ingredient is the morning planning rather than the notes, then the planning is available without any of that.

So the version I kept is a two-minute thing before leaving: what am I going to buy today, and roughly what will it come to. Not a budget, not a limit, just a stated expectation. On days I do it, my spending resembles the cash fortnight. On days I forget, it resembles the baseline. I have been doing this for about four months and the effect has held, though I would say honestly that the size of it has shrunk somewhat — maybe eleven percent rather than eighteen — which I suspect is the novelty wearing off.

The second thing the fortnight taught me is not about money at all. Handling cash made me notice how many of my purchases were made without a decision, in the sense that I could not identify a moment of choosing. Tapping a card takes under a second and does not create a memory. Counting out coins does, and the number of small transactions I could recall at the end of each day went from a handful to nearly all of them. Whatever else it did, it made the day visible.

If someone wanted to try this, I would suggest doing it for the observation rather than the saving, and I would suggest keeping a note of each purchase as you go, because you will not remember and the record is the valuable part. Two weeks is enough. A month would produce better data and I do not think most people would finish it, and an experiment you abandon in week three teaches you nothing except that you abandoned it.

This is what worked for me and I have no idea whether the same mechanism drives anyone else’s spending. What I would take from it generally is that when a technique works, it is worth asking which part of it worked, because the answer is often something cheaper and more portable than the technique.