Just came across a post on Richard Murphy's blog (via Bishop Hill and Tim Worstall, who have both been laughing at a more recent contribution from him) that claimed to show that cutting public-sector jobs would cost the government more money than it would save. For him, this means that we should increase public-sector employment to deal with the government debt problem. I have posted a comment, but I don't know if it will get moderated away, so I thought I would stick it up on here as well, just in case...
You appear to have demonstrated that welfare provisions are more generous than we can afford, and that the combination of tax and withdrawal of benefits creates a very high effective rate of tax on those just above the mean earnings level.
By your maths, someone with gross pay of £25,000 has a disposable income that is £7,365 higher than someone who is earning nothing. That is an effective tax-rate of over 70%. Not great reward for the effort involved. No wonder public-sector employees don’t cost us much more than the unemployed - but is that something to write home about? What you are saying is that we don’t reward people doing work much more highly than people doing nothing.
Your conclusions do not necessarily follow from this calculation. One might equally conclude that we need to rebalance our tax and welfare system so that effective and marginal rates of tax are lower and people are better rewarded for work.