RECENTLY, when some political parties asked the Fiji Bureau of Statistics for ethnic breakdowns of the Fiji 2017 Population and Housing Census results, the government statistician (GS), Epeli Waqavonovono, gave what was to me, an alarming answer: “We did ask about people’s ethnicity, but the decision to release it has to come from higher authority. Once we are instructed then we will release it.”
How many people in Fiji recognise that these simple six words “Once we are instructed” by “higher authority” from Fiji’s government statistician imply:
(a) A potentially dangerous departure from FBS practice, of more than a hundred years of releasing ethnically disaggregated data?
(b) That the FBS and its statistics may now be subject to political decisions by the Government?
(c) That this may undermine the statistical integrity of Fiji’s Bureau of Statistics which has always been supposed to be totally independent of the government of the day; and
(d) The FBS may be in danger of losing the community’s trust in it as an objective provider of national statistics.
Need for neutral statistics
Good bureaus of statistics the world over declare explicitly their belief and practice in the neutral and non-political nature of their statistics gathering, analysis and dissemination.
The Australian Bureau of Statistic (ABS) says explicitly on their website that their purpose is to “deliver relevant, trusted, objective data, statistics and insights”.
The ABS website declares their role as Australia’s national statistical agency is among others, to provide “trusted official statistics on a wide range of economic, social, population and environmental matters of importance to Australia” and “developing and maintaining community trust”.
Neutral national statistics, both economic and social, are absolutely necessary for society to assess in an evidence-based way, where the country is headed, economically and socially.
Society needs data not just on economic growth and distribution but also a wide range of social statistics so as to measure the changes in wellbeing of all sections of the society, including the poorest and marginalied.
While ideally it should be society as a whole which should be assessed, including the private sector and non-profit organisations, in practice, it is usually governments whose performance is most often assessed.
Bureaus must therefore explicitly state their independence of governments, as part of their mission and vision statements.
But all the FBS website states is “with evidence based planning now a requirement for development actors, we as the organisation providing the evidence will continue to face major challenges”.
Note that neutral and objective bureaus of statistics should not be instructed by the “government of the day” to stop producing statistics they have been producing for more than a hundred years, without discussion by parliament.
Departure from
previous practice
Fiji has been conducting censuses since 1881, more than 135 years ago, and every census has reported the population data disaggregated by ethnicity.
Of course, ethnicity was always a political factor since historically different ethnic groups tend to support ethnicity-based political parties looking essentially after ethnic interests.
These groupings blurred at the edges now and then, sometimes having an impact on the election of different governments and that in turn led to military coups in 1987 and 2000.
The 2014 election was unique in that the Bainimarama Government came in with broad ethnic support but it was a government whose prime minister was a military commander still having influence over the military and totally supported by the military for many reasons.
Ethnicity important
for vital statistics
While the political implications may be important to some, ethnic differentiation in censuses is far more important to social analysts because of the other essential demographic information they provide which cannot be made sense of if aggregated nationally.
Ethnically disaggregated statistics on fertility, infant mortality, elderly mortality, life expectancies, child dependency ratios, elderly dependency ratios, rates of poverty, to name just a few, are absolutely vital for assessment of national wellbeing and development, and especially to the Ministry of Health and Ministry of Women and Social Welfare, which is why another name for such statistics is “vital statistics”.
The important point to note is that national aggregates can easily hide what is happening to different ethnic groups.
For example, the annual growth rate between 2007 to 2017 was given by the FBS as 0.6 per cent.
But this is the combined result of indigenous Fijian growth rates (probably around plus 1.6 per cent), growth rates for Fijians of Inidna descent (probably around negative 0.7 per cent and “Others” which has been around 1 per cent in previous census results, but could be anything for the 2017 census.
There are similar profound differences in all the other statistics.
For example, iTaukei child dependency ratios (the ratio of the (0 to 14) age group to (15 to 64) age group) are around twice that of Fijian of Indian descent child dependency ratios, of great importance when one looks at relative financial burdens on families.
Old age dependency ratios (ratio of the over-64 age group to the (15 to 64) age group were about the same for iTaukei and Fijians of Indian descent in the 2007 census (around 7 per cent). But my population projections indicate that while this ratio will have risen for iTaukei to around 9 per cent in 2017, for Fijians of Indian descent it will be a much higher 12 per cent, and projected to double to a massive 23 per cent by 2027.
Wow. What a challenge over the next 10 years for Government and specifically for social organisations serving Fijians of Indian descent whose “mission” it is to serve their communities by providing for age care homes for the elderly where they are comfortable with their diets and cultural practices for religion.
But to get the accurate projections needed for national planning, demographers (and there are still a couple of this rare breed around in Fiji but staying very quiet on the issue) need accurate ethnic disaggregations by age groups.
Why “instruction
from higher authority”
It is unfortunate therefore that the government statistician is stating that he cannot provide ethnic numbers for the 2017 census unless he is “instructed” to do this by higher authorities.
This has never happened before. When did the GS start requiring “instruction from higher authority”? Who would have issued such instructions? This needs to be clarified by the permanent secretary responsible for the FBS.
Does being “all
Fijians” matter?
In replying to the ethnicity-related criticisms by SODELPA leader Sitiveni Rabuka, the GS stated that “On ethnicity information, we are guided by Chapter 1, Section 5 of the Fiji Constitution that states that all citizens of Fiji shall be known as Fijians.” (FT 25 January, 2018).
This might of course hint at where the “instruction” to not issue census statistics by ethnicity, may have come from (even if does not make much sense).
All citizens of Australia may be called Australian, but the ABS still collects census statistics on Aboriginal people, those of Asian origin etc.
All US citizens may be called Americans but they will still collect census information on American Indians or the First Peoples.
Note that it is an exercise in futility for some “higher authority” to issue instructions to the FBS to not issue census data (including potential numbers of voters) by ethnicity.
Any demographer can use previous census figures and with solid population projections, work out what the ethnic voter numbers and proportions will roughly be for the 2018 election.
I estimate the iTaukei will be around 59 per cent of total eligible voters (up from 53 per cent in 2006), Fijians of Indian descent about 37 per cent (down from 42 per cent in 2006) and Others about 4 per cent (about the same in 2006), give or take 1 percentage point for each group.
If the “Others” group is lower, the other two will be higher slightly; and conversely.
But I doubt very much if knowing the number of ethnic voters more accurately will make much difference to the 2018 elections.
* The views expressed are the author’s and not of this newspaper.
