This table has obvious typos. You also omit that the current (LFS) headline measure of unemployment starts in 1971 & BoE make a crude adjustment to older admin data to get its supposedly consistent series. Claimant count shows two govts since WWII saw fall - Heath (1970-74) & Blair/Brown (1997-2010)

We've also gone from headline unemployment coming from admin data (claimant count) to survey based (LFS) which is only from 1971 - BoE do crude revision to older admin data to give impression of consistent series. Using claimants throughout, two govts since WWII have seen fall - Heath & Blair/Brown

You presumably think 66bn too much. Is 52bn? 36bn? Are these stats even comparable? Guess we'll never know if the Economist, like everyone else, happy to outsource its views to the IFS whose 'analysis' is to assume that no area of spending can rise without immediately being in need of reform

yes, this is spend as % of GDP since 1970s. With cuts (pre U-turn) would be falling over this parliament. Govt/IFS etc only focus where spend rising & miss that past data needs revising up for consistency with UC. Also some of rise in incap/PIP reflects more eligible pple claiming - a good thing!

Here's caseload as % of pop by client group, with revision to past inactive to partly adjust for UC reporting changes. Total share flat (lowest for over 20 yrs), inactive rising but still affected by UC reporting. Other takes misleading as ignore breaks entirely &/or only report bits going up

You keep saying it but welfare caseload not surging. Roll-out of UC causing big discontinuities - measured on same basis earlier periods would be revised up. Rise in PIP likely linked to this too - the different way UC works helping cut number of eligible non-claimants, which is a good thing