The modern PhD in the UK is usually dated to the years immediately after the First World War. It is a relative latecomer then, and the complex articulation of a three-cycle process (undergraduate, masters, doctorate) through the Bologna Process is even more recent, and dates to the end of the 20th century.
For a qualification which is so steeped in the appearance of prestige, the doctorate is remarkably under-examined. It has changed relatively little in terms of output, form and process but the context is markedly different.
In 2020/21, there were about 18,000 arts and humanities doctoral candidates. The gap between supply and demand is inverse to the situation in parts of science, where jobs requiring a PhD exceed the supply.
What we have long argued, especially when justifying efforts to recruit, is that the PhD is a route to many careers other than an academic job. And it is undoubtedly true that successful PhD candidates do indeed go on to do a variety of things. The question which is much trickier is the extent to which the PhD was a necessary step, or an elective one which has marginal added value. I will need to have a relevant PhD to enter an R&D role in many private laboratories. I may not need a PhD to enter the civil service, school teaching, publishing, business, a think tank and so forth.