Makes sense
When you consider a local authority paid over £2m for a "SSO" portal for schools back in the BSF days - it was a webpage which authenticated against AD (just one AD, as the entire LA was on a single domain), and then allowed children to click on Gmail or some third party homework site, and Moodle to log in to them. Never worked properly either. Simplesaml.php and a couple of hours does the same thing.
Throw in the modern cloud way of doing things and there's not even a hardware outlay for such a database - you can slap it in AWS or Azure and costs scale with usage - I doubt it'd cost more than a few hundred quid a month in actual resources.
So what they really spend the money on is a team of consultants who will spend 6 months in stakeholder meetings trying to draw up a design spec. The design spec will be absurd and contain things that are entirely unnecessary and somewhat impossible. As they're 3rd party, and get paid by the day, they'll just go with the flow and keep working with the stakeholders. Eventually, they'll finish that stage and hand it over to the actual dev team, who will do their best to build it to that spec and will end up going back and forth with the consultants who will go back and forth with the stakeholders (gotta have the middle-men), because they can't agree what colour they want the submit button to be.
Easy to spend £4m when you do that.
Can't figure out the running cost though.
Biting the hand that feeds IT
