I have been following through the logic of the IFS proposal that VAT be applied at standard rate to all supplies of goods and services apart from new houses (even they balk at that) in the UK.

I have estimated that VAT increases on food will cost £700 a year, on fuel at least £150 a year and children's clothing £100 a year. In median households all are probably an issue. That's about £950 a year.

Median households in the survey data I've used are those who are grouped into income quintile groups i.e. there are five such groups and each has equal number of people in them. The quintiles look like this:

The IFS has done its work on deciles: the data sources are largely compatible though so we can compare one with another. The IFS never mentions value data in their graphs, just decile bands. The published two graphs of the impact of their proposal. The first is that:

This shows who loses what from imposing VAT on all goods and services. the blue line is cash, the dotted line percentage impact.

Note that the impact is horribly regressive: the data shows that.

Note too that at median income levels (bands 5 and 6 here) the cost of charging VAT is estimated to be about £17 a week: a figure very similar to that I calculate if children's clothing is ignored. Gross income of bands 5 and 6 is, on average, about £28,000. No one can argue that is high. But they are losers under this scheme.

Of course, the IFS suggest changes in the benefit system to compensate the poorest for these changes. These would increase benefits by about 15%. With this taken into account the impact is as follows:

The change now benefits those on the lowest three deciles, and is neutral for the fourth. But let's be clear: these are households with total income of about £20,000 or less. I've no problem in protecting them, but the IFS presentation makes it seem as though a loss from decile five onwards is seemingly fair. Quite clearly it is not. They actually lose in cash terms as much as those in deciles 9 and 10 where average household income £73,000.

The evidence is clear, this is very regressive.

If the entire proceeds were put into redistribution to the lower income brackets there would be a logic in what is proposed. Bet let's also be clear, the IFS proposes that as much again as has been put into the benefit package goes into tax cuts. These always have greatest benefit for the well off: it's almost impossible to design a scheme where this does not happen.

If this were to be the case the IFS proposal hits middle England hardest without a shadow of a doubt: the benefit for the wealthiest could far exceed any benefit for the poorest and the person working to make ends meet would pay.

Who in their right minds could come up with a proposal like that?