Making sense of the Census
While digitising data for ease of access and sharing has long been established as a good and useful thing, digitising the collection of said data can often be fraught with difficulties as a model that worked well in an analog world might not make the transition to digital quite so smoothly.
The latest census appears to be suffering at least in part from poor turn out to what is both a legal requirement and also a right of passage that the vast majority of New Zealanders have long been willing to take part in, at least when it was a paper-based system.
But last year's census was the first to be conducted primarily online and Statistics NZ has discovered the hard way that incomplete forms are the bane of many an online surveyor's life.
The Chief Statistician has revealed the figures to parliament for the first time and they aren't too good.
Almost 700,000 New Zealanders didn't complete the forms - including almost quarter of a million partially complete submissions - putting the validity of the 2018 census in doubt.
As politicians on all sides of the house fight over whether anyone can be blamed and whether political points can be scored over the result, the fall out has a practical impact for those who rely on the data to help inform funding models for health and education among other things, and even our electoral system.
While a 90% turnout rate should, on the surface, give us a fair approximation of life in New Zealand, the danger is that those on the periphery are least likely to have filled out the forms and are most at risk of being overlooked. Having said that, the previous census had a completion rate of 94.5% so the drop isn't precipitous but it does mean we have to rethink what happens from here in terms of census data.
Are bulk online forms the way to go for the next census? Should we revert to a paper-based system or soldier on with an online model? Do we need to increase the fines and number of prosecutions for failing to complete the census forms fully (the stick approach) or should we spend more on marketing and advertising the census in the lead-up to the next one (let's call that the carrot for want of a better term)?
It's a conundrum that the people at StatsNZ will have to deal with and is something that vexes most pollsters in both commercial and political senses of the word.
You must be logged in in order to post comments. Log In