Taking a closer look at Labour’s capital gains tax

Sunday, 2 November 2025

Vernon Small is a journalist and former Labour Government advisor.

OPINION: First, the good news from Labour’s capital gains tax announcement.

In promising to implement an expanded CGT, albeit a timid one, the party has done (some of) what every major economic agency in the world and at home has said is needed.

It is crucial that we broaden our tax base beyond the current mix of consumption, business and labour taxes to pay for the increasing demand for spending on health, pensions and infrastructure, especially as the proportion of working age people in the country shrinks.

We can’t just cut our way out of that dilemma. It’s the critics, who say a CGT would wreck the economy and bring devastation to businesses across the land (in contrast to almost all the other developed countries of the world that have one), who are out of touch.

Of course, it is not the first time Labour has bitten the bullet on promoting a capital gains tax. Nor is it the first time the party managed to shoot the same bullet into its own foot.

This time it was courtesy of a leak to the media just days before the policy was to be made public, causing Monday’s scramble to announce it.

Presumably the leaker was miffed the party had opted for such a limited CGT, taxing only commercial and residential properties (excluding farms and the family home), and had turned away from a wealth tax which is popular with many rank and file members.

It is a fair point that a CGT – as opposed to a wealth tax – targets second and third tier wealthy rather than the top 1-3%, would raise fewer dollars early on and would do less to span the abyss between the tax paid by wage and salary earners and by the mega-wealthy.

But Labour leader Chris Hipkins and finance spokesperson Barbara Edmonds give greater weight to fears a wealth tax would see capital flight, especially to Australia, that could cost jobs here.

For now, that internal debate is water under the bridge for Labour. So, whatever the motive for the leak, it was counterproductive and attack-line gold for the Government.

From that quarter we have also seen the expected mix of marginal cases and supposed unfairnesses trotted out. They include that the tax would hit the inflation component of capital gains (as does tax on the interest on savings and on the inflation-compensating component of wage increases) and even a claim it would undermine KiwiSaver accounts because they are invested in businesses that would pay CGT on commercial property profits.

It all looks a bit desperate, especially coming from National that over the years has made more policy changes that undermine the future value of KiwiSaver accounts than any other party. There are, however, some rough edges to smooth, including the treatment of the sole family home that has, for whatever reason, not been lived in.

Chris Hipkins, Ayesha Verrall and Barbara Edmonds did a steady job fronting the announcement, despite the rush after the leak. Edmonds was particularly impressive handling detailed queries … as you should expect from a tax specialist. Her role will be crucial in deflecting attacks at the “what about” level.

But Labour will need to be careful that the whole of its front bench, particularly Hipkins, is across the brief. It was the hailstorm of micro-attacks, and previous leaders’ inability to rebut them, that scuppered earlier attempts at a CGT as much as hostility to the tax itself.

Much depends on the projected revenues from the flat 28% CGT.

Some have suggested Labour’s numbers are over-egged and heroic, especially in the light of the soft housing market.

But the forecasts assume house price growth of 3% a year and that is not out of whack with official and other forecasts. Conservative almost.

The Reserve Bank expects house prices to increase by 3.9% in 2006 and 5% in 2027. Treasury has a more modest 3% by 2027. Westpac and ANZ pick 5% or more in 2006, before Labour’s policy would kick in.

Linking the CGT to spending on health, specifically three free visits to the GP a year for all, was sensible politics.

But in truth the link and the “Medicard” that gives the promise tactile form come with a fair amount of haze and tricks of the light.

The three visits can’t be accumulated to subsequent years if they are not used, can’t be used at another GP practice and the cash is paid to the GP practice whether the visits happen or not. So, it is more accurate (but less politically sexy) to see it as an increase in GP funding for the first few visits.

Then there are the cost and revenue implications.

The free visits policy, including the card and some other elements, is estimated to cost about $550m a year.

The CGT revenue – which Labour has said will be ring-fenced for health spending - is forecast at $100m in the first year (2027-28) rising to $1350m in 2030 and the years beyond.

That’s an average of $700m over the four-year forecast period, leaving an average “freeboard” of $150m or more a year. Labour used that average figure to finesse away the significance of a $300m to $400m shortfall between revenue and costs in the first two years.

More interesting is what happens from 2030 onwards when the revenue – if forecast correctly – would exceed the costs by some $800m a year.

Labour has said all the CGT money would be ring-fenced for health – essentially a hypothecated tax – but it is not clear if that means spending it all on new health initiatives or using some to offset other health funding needs, such as from an aging population.

It may seem like splitting fiscal hairs, and to some extent it is. But it would have a significant impact on Labour’s overall health spending track.

That should become clearer next spring when Labour releases its full fiscal plan, alongside some other key decisions; such as whether to reinstate depreciation for commercial buildings and allow interest deductibility for landlords to offset CGT.

But for now, at least we have a contest of tax ideas.

What do you think? Email sundayletters@stuff.co.nz. Please include your full name and address.