Ms Deirdre Donaghy:

I can give a bit more of an explanation as to how certain figures come up that might explain some of the difficulty that would arise in trying to do that. Income tax is the easiest context in which to give an example; in order to cost any measure that is proposed for any given year, Revenue does a lot of work to construct a base for it and in general that is based on tax returns, which is the best data that can be had because it is population data for the country. Tax returns are filed after the fact, however, so they are always a year or so behind. For example, for 2019 there will be a 2016 base. That means that data from 2016 will be used to give comparative full population data for 2019 and that data is then grown up to reflect the 2019 tax base by applying macros to account for increases in employment and income and then that is used to try to estimate the cost of measures. The difficulty with that is that a lot of assumptions are being used and that is the only way it can be done because I cannot tell the committee today exactly what will happen tomorrow; I can only estimate it and it takes an awful lot of work to do those estimates. The question is whether that is the best use of time or whether we would be better served doing something like a cost benefit analysis where we can look at what has actually happened, albeit that is what actually happened last year, and then use that to make the policy decisions going forward. That might explain the technical difficulties that would be involved in trying to project what we expect the costs of any particular tax measure for next year to be.