Statistics at dawn: LA vs WAO

Post by @TobyMasonBBC

Another fractious day ahead in another fractious political week it seems. This morning, the former education minister Leighton Andrews and the Wales Audit Office have taken to the airwaves to debate whether the Welsh Government's flagship student fees policy has come in millions of pounds under or over budget.

It's pretty unusual for the WAO to face this kind of questioning about their reports which are by and large treated as gospel even by those who are criticised.

But the response shows just what a political hot potato this policy now is, and how high the stakes.

Here's an explainer on what this disagreement is about: as you might expect there are a fair few numbers.

The WAO have based their analysis of there being a 24 per cent increase in the five year cost of the policy from the initial £653m estimate to £809m currently on the core assumption used to underpin the Cabinet's decision to pursue the policy in November 2010.

This assumption at that point in time was from academic year (AY) 2012/13 that there would be a maximum variable tuition fee of £7,000pa at Welsh and English HEIs. This £7k model works out the annual cost of the tuition fee grant as:

2012-13: £31.5m

2013-14: £92.7m,

2014-15: 147.3m

2015-16: 182.9m

2016-17: £198.9m

= £653m

The reason the WAO say they have taken these figures as their starting point is that in order to audit effectively you have to have a baseline to start from when you're evaluating the cost of a policy decision.

And since the £7k model was the main thrust of the 2010 Cabinet paper seeking approval for the policy, that is what they have gone with. There was a reference in that paper to the financial risks to the policy of a £9k maximum fee level - but it's not believed that this reference mentioned that those costs had been modelled and the actual costs were not provided.

This was confirmed in general terms by Mr Andrews on Good Morning Wales earlier, within the bounds of the convention that former ministers don't speak in detail about private discussions while they were in office.

There's no way of us checking this independently as the paper is still restricted.

However, at the same time that the officials were doing the detailed modelling on the £7k assumption, they were also modelling the cost impact of £9k - this gives a total cost figure over five years of around £1.02bn. Both of these were subsequently published in 2011 following BBC Wales FOI requests.

Leighton Andrews' argument is that since that £9k modelling had taken place within government at that stage, then the WAO are wrong to take the costs of the £7k model alone as the starting point for their cost inflation analysis - since the WG did have higher estimates in its possession. This, as he has pointed out robustly this morning, is acknowledged at section 1.43 of the WAO report (or "buried" as he described it).

But as the report makes clear, these £9k cost risks weren't shared with the full Cabinet and the paper before them didn't rely on them as the basis for their decision. The WAO say that "certain" Cabinet members were aware of the £1bn+ cost risk of the £9k model - Plaid Cymru sources say they are fairly sure the then deputy first minister and Plaid leader Ieuan Wyn Jones was made aware of it.

Subsequent to the November 2010 Cabinet decision and announcement much clearer information became available as to the exact fees that HE institutions would be charging and therefore the £9k model was far more accurate - so then at that point the estimated cost of the policy became £1.02bn over five years.

Since this forecast was made, the expected cost of the tuition fee grant has decreased from £1.02bn over five years to £809m.

One of the main reasons for the decrease in the forecast cost of the TFG is that the average tuition fee for Welsh students has decreased from £9,000pa to £8,680pa for 2012-13 and from £9,000pa to £8,291pa for future years.

So the reduction is due to the new £9k estimate being the absolute highest possible amount the policy could cost and it actually coming in a bit lower than that - hence the fact that Mr Andrews and the WAO are about £350m or more apart in their interpretations.

If nothing else, it's a fascinating story of the inner workings of the Cabinet, and the way in which decisions with vast political and cost implications are taken.