The problem with this statement is not that it is not true, because never a truer word was spoken, but rather that the statement is not featured in road safety advertising or enforcement to the extent that its life saving potential truly demands.

The two graphs below illustrate how dramatically road deaths increase with small increases in the percentage of people not wearing seatbelts. The reason there are two graphs is because it is only possible to easily do this comparison for front seat occupants. From the differences between the two graphs it would appear that non-speed related fatal crashes in the more tourist oriented regions involve a much higher proportion of rear seat occupants and therefore the front seatbelt wearing rate is less important than in the other regions. The relationship between seatbelts wearing and lower death rates is extremely strong for all regions but it is a nearly perfect cause and effect relationship for the non-tourist regions.

Sources:

Non speed related deaths - Shattering illusions: saving lives, LTSA, 1997

Seat belt survey results - Motor Accidents in New Zealand, LTSA, 1998

Vehicle km travelled - National Traffic Database, Ministry of Transport, 1996