The thesis goes like this: sack a council, quango or government worker and the economic costs are fourfold. There is a strain on the Exchequer because the former employee pays less tax; he or she requires some sort of pay-off in the form of a redundancy payment or whatever; the former employee is then no longer providing a service to the public thus reducing output and quality; the strain on their health due to forced retirement and reduced income places a further burden on local and national services.
When you look at the costs of all that then you might as well keep people on the payroll. This analysis is partly flawed and outdated. What should concern decision makers is the level of productivity within the public sector. That level is less than it should be because of procurement policies, levels of absenteeism, interdepartmental wrangling and an imbalance between the numbers of managerial and front-line staff.
Improving service to the customer may mean getting rid of staff. But there ought to be a differential between between people's post-tax incomes after they've gone and what they earned while in work. That may mean further pension and welfare reform. There does need to be a recognition that many are happier doing voluntary work they enjoy rather than paid work they find stressful.
In the modern world there is no case for public sector over-staffing simply because the costs of reduction are too high. Is that job necessary and can performance be continuously improved? That should always be the question at the front of decision makers' minds.