Pacific statistics: Census of a small island nation
While the Australia Bureau of Statistics (ABS) readies the release of 2016 Census information, the tiny Pacific Island territory of Tokelau have been there and done that!
With the help of Statistics New Zealand, Tokelau undertook their Census in October 2016 and have recently released the full results. Speedy huh!?
Life is good on Tokelau (population: 1,197)
Tokelauans are satisfied people! Around half those who responded to the 2016 Census rated their life satisfaction 10/10 (49.9% of residents). The next most common response was 8/10, selected by 36.1%. Only 2.5% of the population gave a rating of 4 or lower.
Cross-tabulations found that older age cohorts reported higher levels of satisfaction, as did those that felt they were in good health.
Tokelauans also reported a powerful sense of belonging, with 58.6% giving a 10 rating. Interestingly, residents of the Tokelauan atoll of Fakaofo felt this sense of belonging very strongly, with over two-thirds of the adult population (67.8%) rating sense of belonging as a 10/10.
New Zealand undertakes a General Social Survey (GSS) every two years and asks the same question on life satisfaction. In 2014 only 17.8% Kiwi respondents rated their life satisfaction as 10 out of 10. By comparison, Australians rate their overall life satisfaction as 7.6 out of 10, according to this general social survey, conducted by the ABS in 2014.
The Tokelau Census is incredibly important, as it is the only comprehensive source of demographic and economic data for the three atolls (Atafu, Fakaofo, and Nukuonu) that make up the island nation. They now have data from three Censuses (2006, 2011, and 2016).
Not all Censuses are created equal though, and the Tokelau Census is definitely not a cookie-cutter version of New Zealand’s. In 2016, for the first time, Tokelau also asked questions on the quality of life including reflections on health, sense of belonging, social connectedness and income adequacy. The power of a Census-style approach in capturing these wider indicators of community wellbeing cannot be underestimated.
Making the move to electronic collection
Tokelau also changed their collection process, embracing technology in 2016 by committing to a tablet-assisted questionnaire using a computer-assisted personal interview approach (granted, Tokelau only has a small population – a total of 1,197 usual residents who were present in Tokelau on Census night, and another 302 usual residents were away overseas).
With limited internet access and the additional training required, the decision to switch to an electronic collection process wasn’t an easy move to make. However, the Tokelau administrators reported a handsome payback; the CAPI format was user-friendly, and the self-correcting function of the technology meant that many mistakes were picked up and resolved during data-capture, making the post-enumeration phase much more efficient.
So what did the 2016 Tokelau Census find?
- The usual resident population count was up 4.7% on the 2011 result (1,197 compared to 1,143)
- Tokelau has a young population (albeit getting older each year). In 2016, the median age (URP) for Tokelau was 25 years. In 2011 it was 24 years, and in 2006 it was 22 years.
- The fertility rate is 3.8 children per woman, down on 2011 (and lower than neighbouring Samoa)
- The population experiences huge migrational movement – most Tokelauans (59.2%) have lived overseas for more than six months at some time in their life (the 2013 New Zealand Census recorded over 7,000 residents identifying as Tokelauan descendants).
- Religion is a central part of Tokelauan life with most people being Christian, though religious denominations (congregational Christian, roman catholic or Presbyterian) vary hugely between atolls.
- More than half the adults smoked (51.3%) up slightly on 2011
For more fascinating insights into our Pacific Island neighbours, you can access the Tokelau Census results here.
.id is a team of demographers, population forecasters, spatial planners, urban economists, and data experts who use a unique combination of online tools and consulting to help governments and organisations understand their local areas. Access our free demographic resources here
How do they know that exactly 302 residents were overseas on census night. For that matter how do we in Oz know the comparable number for us (especially at sub-national level) and indeed in the absence of anything like an administrative census how do we know how many residents there are for the purposes of ERP which obviously is a better number than census enumerated residents?
Hi Ian,
yep good questions. In the case of Tokelau, their report (included in the blog link) recognised that the movement of their people presented a unique challenge in counting the population. Their report notes ..
” In the 2016 Census, approximately 20 percent of the usually resident population was away from Tokelau. While Tokelauans travel overseas for many reasons, healthcare and education are two notable reasons for absence. Due to the high number of people absent from Tokelau on census day, in 2006 we developed stringent criteria to determine who was a ‘usual resident’. These criteria were applied consistently across the 2006, 2011, and 2016 Censuses… Demographic information was completed by the head of the household for usual residents who were away on census day (absentees). Where the whole household was absent on census day, the census day supervisor for each atoll was responsible for answering basic demographic information for the household.”
The Tokelauan’s called this count the de jure usual resident count. In 2016 the de jure count of 1,499 was made up of 1,197 usual residents who were present in Tokelau on census night, and 302 usual residents who were overseas. The absentee sub-population of 302 included 48 Tokelauan TPS employees and their immediate families based in Apia, and 254 usual Tokelauan residents who were overseas at the time of the
census.
Tokelau has a tiny population so the census collectors/supervisors for each of the three Atolls would feel confident completing the census form for an absent household. Clearly the same approach could not be taken in Australia or New Zealand as census supervisors would not have the same intimate knowledge of local populations. In New Zealand we rely on our population filling out the census forms if they are away from their usual place of residence.
Yet the usual place of residence is an inherently tricky concept. For example there is no time criterion given for when a place becomes a usual residence, largely because the concept is self-determined and associated with perceptions of belonging. Stats NZ provides some interesting notes on defining usual residences …
“If you are an overseas resident and will be staying in New Zealand for less than 12 months, give your address in your home country. Otherwise, give your New Zealand address.
If you are a New Zealand resident, follow these guidelines to give the right address.
If you are a primary or secondary school student at boarding school, give your home address.
If you are a tertiary student, give the address where you live during the semester.
If you live in more than one dwelling, give the address of the one you most consider to be your home. If you spend equal amounts of time at different addresses, give only one of those addresses.
Children in shared care should give the address where they spend most nights. If children spend equal amounts of time at different addresses, give the address of where they are staying tonight. (see http://www.stats.govt.nz/methods/classifications-and-standards/classification-related-stats-standards/usual-residence/definition.aspx )
Finally, your note picks up on the ERP (Estimated Resident Population) as a better number than the census enumerated residents. The ERP is certainly the official population figure however the URP data is still used for understanding the socio-demographic characteristics of a population.
Glenn Capuano, our resident census expert talks to this subject in his February blog “How should I use enumerated data usual residence data or estimated resident population figures?”
See http://blog.id.com.au/2017/how-to/should-i-use-enumerated-data-usual-residence-data-or-estimated-resident-population/
a very helpful answer, thank you and I wish that some in local government understood the data better. Perhaps local governments should be invited to assist with post-enumeration surveys.
of course, definitions (and data) are only good if respondents take them seriously, for example with increasing online completion of census forms and more than 2½% of residents as defined being overseas on census night (on the basis of time away reported in arrivals and departures statistics) how many residents out of Oz or NZ on census night completed their forms anyway? These numbers could distort both counts of residents and socio-economic data on resident populations in local area with populations prone to travel (as indeed in the case of the Tokelaus)