Tauranga artist Timo Rinnali is among those resisting a perceived voter swing in favour of a new flag.

Change the NZ Flag Chairman Lewis Holden says today support for changing the flag is increasing, while support for the existing one is softening.

Timo Rinnali is supporting the current flag.

A poll of 1000 people earlier this month showed 36 per cent of people support changing the flag. Last September support for a new flag was 25 per cent of those polled. Support for the old flag has dropped from 69 per cent to 56 per cent since September 2015.

'This shows New Zealanders are starting to embrace the new flag design because it's more representative of our nation,” says Lewis.

Timo, who submitted his own alternative design for a new flag, remains unswayed.

'This morning, and I might change my mind tomorrow - but this morning I'm thinking I wouldn't change the flag to the one they have chosen as the opposition.

'My problem with the new design, is it's a design. It doesn't tell a story in my opinion. I feel a little bit like perhaps the public were bulldozed into that particular design.”

If there was to be a new flag the new one needed an image that has some connections, like a story board, and without that connection no amount of fancy designs would work.

'I still believe that. I don't know if the new one does it for me, that's all. If there was another design I really believed in, I possibly would have voted for a new flag.

'John Key really promoted that design and really pushed for it. And I think it almost seems that way in my opinion, so it was like a foregone conclusion.

'This morning I'm feeling that I wouldn't vote for change with that particular design. I'm not saying it's wrong to have a new flag but that particular design – I think I prefer the old one.”

Tauranga RSA members are also continuing to oppose the proposed change to the New Zealand flag, says Tauranga RSA general manger Graham Howard.

Commenting on claims that opposition to the flag change is softening, Graham says the membership voted on the stance to take in the flag debate at the annual meeting late last year.

'We asked members at the AGM at the end of last year what our stance would be and the membership voted on it and the stance was keep the flag the same.”

The vote at the national council on the issue was the same, says Graham.

In the first referendum, over 1.1 million New Zealanders voted for the two fern designs and that number is expected to grow in the second referendum next month.

'In our view, a higher turnout in the postal ballot will improve the chance for change.”

A Curia Research poll from 28th January to 2nd February shows support for the new design is growing, says Lewis.

'People are thinking carefully about what's best for New Zealand and there's a mood for change emerging. Feedback shows the main reason for change is preference for the new design and differentiation from Australia's flag.”

The Change the NZ Flag initiative is working on an online campaign to promote the merits of the new flag and encourage people to vote for change.

'We've got some exciting developments over the coming days and weeks, where people will see the benefits of a unique New Zealand identity that's different to Australia's and any other flag.

'We have an opportunity to stamp our own mark on the world, while also respecting our rich and proud heritage. The new flag provides that opportunity.”

The second binding referendum on the New Zealand flag will be held from 3 March to 24 March 2016. New Zealanders will be asked to choose between the current flag and the preferred alternative design selected in the first referendum. The flag that receives the most votes in the second referendum will be the official flag of New Zealand.

These will be postal referendums, so voting papers will be sent in the mail. Only those who are enrolled before voting starts are able to vote in the referendum.

Curia Research conducted the poll of 1000 eligible voters from a selection of 15,000 nationwide phone numbers, from 28 January to Tuesday 2 February 2016. The results are weighted to reflect the overall voting adult population in terms of gender, age, deprivation and area. The maximum sampling error for a result of 50 per cent is +/- 3.2 per cent, at the 95 per cent.