@nearlylovemyusername ,
Thank you for your measured post.
The thing is that the ‘propensity to spend’ is an input number. The modeller didn’t assume 100%, I vaguely remember he assumed 70% but, like all modellers, he looked at lots of different scenarios with both the number leaving lower or higher and the propensity to spend lower and higher. Modelling doesn’t give precise answers but ranges and breakevens. It is still better to have a model than just assume an answer!
I do think that, realistically, people may have very lofty ambitions for the saved money but, ultimately, money is fungible and, for instance, someone who has driven an old banger for years in order to send their children private will, when they realise they have £40k sitting in their bank, decide to upgrade. And, maybe treat their family to a nice holiday . Now, that won’t be all the people but it will definitely be some of them. So, we can argue about the size of the number, but it is non zero.
And some parents, who either wanted a good private or state, may now realise that the private they wanted is looking at lower attaining pupils to fill the places, and may well switch state to private. I have a strong feeling this has been going on for years now, with less well off parents of higher attaining children being substituted by wealthier lowers attaining ones.