The only reason McCreevy's rate cut resulted in increased CGT receipts was because there were large accrued but unrealised gains. People were motivated to realise them by the tax cut. But that had to be a one-off; once accrued gains have been realised and taxed, they can't be realised and taxed a second time.
There are two lessons to draw from this.
- First, a CGT rate cut can only lead to a rise in CGT receipts if there are large accrued but unrealised gains, and current CGT rates are operating to disincitivise realisation. That isn't necessarily always the case; you'd need to be satisfied that those conditions prevail now before cutting the CGT rate.
- Secondly, the rise in CGT receipts will be temporary. All other things being equal, after a year or two you would expect to experience a decline in CGT receipts — you've cleared out the backlog of accrued gains, and newly-accruing gains are being taxed at the new, lower rate. And that's exactly what happened with the McCreevy cuts — a rise in CGT receipts for two or three years, followed by a reduction. (Receipts rose again from 2003 onwards, but that was because of asset price inflation.)
The second point means that there is a long-term tax cost to the exercise. This is, all going well, masked in the early years by the flushing out of accrued gains, but that is only temporary. If you're going to suffer a tax cost for this, you want to make sure that it goes towards acheiving your objective of stimulating more housing construction.
How did the McCreevy CGT cut affect the rate of housing construction? It didn't, basically. The rate of housing construction, having been static for over a decade, increased steadily from 1993 onwards, and that rate of increase persisted until 2003, after which it rose sharply. The CGT rate cut in 1997 doesn't appear to have had any great effect on the rate of housing construction; there's no evidence that the 40% CGT rate constrained housing construction in the years before 1997, or that the reduction to 20% stimulated housing constructin in the years after 1997. If the rate cut had been presented as a measure to stimulate housing construction, it would have been judged a failure.
History suggests, then, that an across-the-board CGT rate cut is not an efficient or effective way of generating more housing. The reason for this is fairly obvious; most of the assets disposed of to take advantage of the rate cut are not development land. A deeper rate cut, but applying only to development land, looks like a more efficient way to target the policy outcome; a rate cut applying only to development land that did in fact get developed would be more efficient still.
(But no CGT rate cut will be effective unless the problem, or one of the problems, is that the owners of development land are being constrained from make it available for development by the current CGT rate. While that's not implausible, you can't just assume that it's true. If there are othe obstacles, and you don't identify and address them, even a targetted CGT rate cut is not going to acheive very much.)