Skip to content

“Fiji’s intriguing 2017 Census numbers” (FT 27/1/2018)


Fiji’s intriguing census numbers (FT 27/1/2018)

How many Fiji Times readers over the last month have thought twice about some very important census numbers produced by our Fiji Bureau of Statistics?

One was the census result that the rural share of Fiji’s population in 2017 was only about 44%, declining significantly from about 49 percent in 2007.

The second more difficult one was the age distribution of Fiji’s population given in five year cohorts, and what it implied for the potential number of voters in Fiji’s 2018 elections.

I have every confidence that these numbers produced by the FBS are reasonably accurate and can be depended upon, contrary to the views of some politicians..

If these census numbers do not gel with other numbers, even some GDP numbers produced by the FBS itself, then the public might want to question the accuracy and significance of these other numbers, not those of the 2017 census.

There may be some interesting answers.

Total Population (2017)

The FBS estimate of the total population in Fiji the night of the Census (17th September 2017) was 884887.

Given that FBS estimated that they covered almost 99.5% of all households in Fiji, they believe that their census total was most accurate.

My own population projections a few years ago estimated that the 2017 population would be about 886,438 or just 1,500 more than the census figure.

Keep in mind that projections depend on assumed emigration numbers, assumed fertility rates and assumed mortality rates remaining as projected, which will naturally not be the case in reality.

Such a small (0.2%) difference from the projection would suggest that the FBS Census total, and their age distribution numbers are pretty accurate.

So what is the big deal, you may ask?

The big deal is the number of potential voters aged 18 and over indicated by the census figures disagrees quite significantly from the total number supposedly registered by the Fiji Elections Office (FEO).

Potential voters indicated by census and FEO

The FBS gave age disaggregated numbers of the total population in Fiji, by five year age groups such as (0-4) (5-9) (10-14) (15-19) (20-24) etc.

Unfortunately, they did not give the numbers for those aged 18 and over (voters) but this can be roughly estimated from the five year numbers given by adding 40 percent of the (15 to 19) group to those 20 and over.

My estimate from the FBS census numbers is that the potential number of voters in 2017 was 580,646 which number the FBS has confirmed to be accurate (personal communication from FBS Government Statistician).

If we then increase this number by the annual growth rate of population (0.6%) we get 590 thousand potential voters for 2018.

BUT the Fiji Elections Office estimate of voters registered at the end of 2017 was a much higher 624 thousand voters, or 34 thousand more than the census estimates.

With only around 7 thousand overseas registered voters, how do we then explain the extra 27 thousand voters registered by the FEO?

Are the FBS numbers inaccurate for some reason we do not know about?

These are questions that ought to be of interest to the Fiji Elections Office, the Fiji Bureau of Statistics, and all the political parties and candidates.

Rural:Urban Drift

The second intriguing number arising from the 2017 Census is that the rural share of the total population was around 44 percent in 2017, down from 49% in 2007.

Some of this decrease in rural population may be the result of the FBS expanding the urban boundaries where cities and town have grown.  But surely not by that much.

The Government and all political parties need to request the FBS to estimate exactly what proportion of the increase in urban share in 2017, was due purely to the urban boundaries being increased and what is due to continued rural:urban drift.

I suspect that the boundary effect will not be more than about 1 percentage point and that there remains a clear trend of rural urban drift worsening 2007 and 2017, usually because of people’s search for jobs, incomes and better standards of living.

I hear some asking: so what is the big deal since this is the long term  trend anyway?

The big deal is that since 2007, the Bainimarama Government has spent enormous amounts of taxpayer funds on rural roads and other infrastructure in both Viti Levu and Vanua Levu, to improve living conditions and livelihoods for rural people, an important development objective for all governments.

Why have these new terribly expensive roads, which have virtually doubled Fiji’s Public Debt,  not reduced the rural:urban drift?

The Ministry of Education has attempted to improve schooling in rural areas and attempted to keep students to their local schools through “zoning” rules.  Have these zoning rules not worked?

Have health services in rural areas not improved enough to keep rural people there?

If these incredibly costly development initiatives by the Bainimarama Government are not working in the rural areas, surely it is incumbent on them and the Opposition Parties to collectively find out why they are not working.

I remind that similar (but perhaps less costly) efforts were also made by previous governments, with similar failures.

Other intriguing FBS data

I have written previously that the 2015-16 Employment and Unemployment Survey results from the FBS indicated that between 2010-11 and 2015-16, the number of persons “working for money” declined by 10,138.


How can this be compatible with claims that Fiji has “enjoyed” 8 years of successive growth rates of GDP and that there has been a decline in the Rate of Unemployment?

Given that “Unemployed” is defined as a person unemployed and “actively looking for work”

have large numbers of people simply given up looking for work?

Another worrying number from the EUS was that those working for money in Agriculture declined in the same period (between 2010 and 2016) by a massive 31 thousand (or by -54%).

There is also disaggregated constant GDP data on the FBS website that shows that agricultural GDP went down in 2016. Why so? Surely not all due to Cyclone Winston?

These are all interesting questions that need to be addressed by academics at our three Fiji universities and our technical civil servants.

There may be some complex answers that reconcile all the apparently paradoxical numbers from the FBS, but they need to be clarified if taxpayers’ funds are not to be wasted.

[One reason suggested by FBS is that the the Census only counted those who were in the country on the night of the census, not those outside of Fiji. Hmmm.]

[For reasons of space, I won’t comment here on the news that the Ministry of Sugar is planning to bring in foreign cane cutters].


Comments are closed.

%d bloggers like this: