There are quite a few things confusing about this, but then we have to keep in mind that this is KPMG survey so they are skewing it in their own interests.
1) Permanent jobs are not really as much of an indicator of seasonal issues as contractor / temp roles:
That's the finding of a new survey, which revealed that according to a seasonally adjusted index measuring permanent vacancies in the IT sector, demand to fill permanent staffer jobs in the IT market had risen to 64.4 per cent in the dog days of August, up from 62.8 per cent a month earlier.
Permanent roles are long term commitment so if this is true, it isnt really seasonal. Its companies deciding they need more employees for something and will continue to need them for a long time.
2) Contractor jobs are an indicator of seasonal trends:
Meanwhile, demand from Blighty companies to hire temporary techies fell to 58.7 per cent from 59.1 per cent in July.
So in August there was less of a need for a temporary work force to fill gaps.
Seems the headline is a bit assbackwards.
There is a point 3 though.
3) This data is likely to be meaningless. In all likelihood, KPMG searched round the job boards and simply counted the adverts. This means that when (like my current role) it is advertised by 12 different agencies - all with slightly different details - it counts as 12 jobs rather than 1. It also means that some non-existent jobs (posted by unscrupulous or clueless recruiters or hiring companies) get counted when in reality there are none.
The easiest way to tell if demand has outstripped supply is to look at the average salaries and contract rates available. These are still, largely, in line with 2009 figures - and as other posters have said include shopping lists of skills for £35-40k a year.
All of this implies to me, at least, that there isnt enough of a skills shortage for anyone to actually care - it is just a shortage of skilled workers prepared to work for the salary they had when they were unskilled.