Political Correctness: Sex, Race, Ethnicity*and the ‘Average’ Fijian
By Crosbie Walsh
LIKE SEX AND AGE, RACE AND ETHNICITY are realities of life, and they are — or should be — included in all statistics where they are relevant. Not including them results in a loss of important information and insights into causes, effects and inter-relationships. But this is what Fiji appears to have done, presumably for new "politically correct" reasons. There appears to be no Fijians, Indo-Indians and Others in Fiji any more: they are all Fiji citizens. Please, someone, tell me I'm wrong.
Removing or not using race or ethnicity will not make race or racism go away. Fiji’s people are no less itaukei, Indo-Fijian, Kai Solomoni, Kailoma, European, Chinese or so on now that all citizens are called Fijians. Calling all citizens Fijians will not eliminate racism, although it should help to remove institutional racism. A healthy Fiji will be one where each race or ethnic group proudly proclaims its unique self and its shared nationality.
Arrival and Departure Statistics
The apparent removal of race from official statistics came to my attention when I asked the Bureau of Statistics for information on arrivals and departures. Despite several requests, all I was provided were tables distinguishing citizens from non-citizens. This means that while previously I would have been able from departure information to explore relationships affecting the so-called brain drain, for example, I now had one less variable to feed into the matrix. All I could examine is the “average” citizen. In some societies this may be sufficient but not in Fiji whose history from colonial times has produced very different demographic (births, deaths, internal and international migration), social (health, education, housing, religion), economic (occupation, industry, income, expenditure, savings) and locational patterns for each of its races. Race is probably the most important single explanatory variable because it allows us to see variation around the average. A government that wishes to monitor progress towards its goals using statistics that are based solely on common citizenship, with no mention of race, is likely to come up with statistics that are next to useless.
The National Average
Citizenship will produce a national average which may be necessary for international comparisons but it is probably the least useful statistic for use within Fiji because it hides differences and extremes. Graphing such averages helps a little because graphs show the most numerous occurrences (the mean) and differences around the mean and average. Often the graph will shown a normal curve that looks like an inverted bell but in Fiji the norm may consist of two extremes (a bipolar distribution) which when merged only look normal. Thus, to talk about average income, average housing, and just about average anything else makes no sense at all. Fiji’s average income is not shared around between rich and poor, and its average house is not a mix of two-storeyed concrete houses with four bedrooms, toilet, shower, lounge and kitchen and one-room corrugated iron squatter shacks. The average migrant is not the difference between gross arrivals and departures.
Rural, provincial and urban averages
Neither can one sensibly talk about a rural or provincial average. Rural areas typically comprise quite different itaukei and Indo-Fijian residential patterns: itaukei live in communal villages; Indo-Fijians live on their individual farms. And this often produces differences in access to roads, electricity, water supply and toilets. Ba province is predominantly Indo-Fijian; Lau province is almost entirely itaukei. The average province is not Ba and Lau combined.
These patterns reproduce themselves in the towns. Some suburbs are predominantly one race or another, but within most suburbs major differences are evident. I used to live in Namadi Heights, an affluent suburb in north Suva. It was once mainly European with some Indo-Fijians. Today it is more racially mixed but across from where I lived there was a settlement of squatters who for census purposes were added to their affluent neighbours. It is for this reason that I think the use of “residential areas” in the Household Income and Expenditure is likely to give inaccurate results. The use of housing types would be far more accurate. I proposed five types that were used and showed their worth in the 1996 census but they have not been used since.
More race sub-categories would be useful
The value of race as a helpful explanatory variable in censuses and surveys has been demonstrated so many times that further advocacy should be unnecessary. Indeed, a strong case can be made to further refine racial categories. The first time I was aware of this was during a household survey of Navua township in the mid 1970s carried out by second year USP students as part of their coursework. We divided race into Fijians, North Indians, South Indians, Gujerati, Muslims and Others. These were not pure categories, of course. A person could be a North Indian Gujerati Muslim but crude as the measure was, it produced far more interesting results than had we kept to the usual Fijian, Indian, Other categories. There were, for example, major differences in household composition, housing, education, place and country of birth, occupation and income which would have been concealed had we treated all Indians as one.
At the regional and national level it would also make sense to treat separately the different races that make up “Other”. A composite of usually affluent Europeans and Chinese; usually quite well off Rotumans , Part-Chinese and Kailoma (part Europeans), and usually poor or destitute Kai Solomoni produces the most meaningless average imaginable. If government wants to do something to help the Kai Solomoni and other descendants of Fiji’s first indentured labours, or obtain information on Rotumans, Tuvaluans, and Banabans in mainland Fiji, a sound set of statistics would be a helpful starting point.
Up until a month ago I had no intention to write anything about retaining statistics on race, largely because I think too many people see Fiji through a racial prism that explains everything that happens politically, economically and socially. But if reference to race is often over-done, it can also be underdone. Race in combination with other useful explanatory variables — such as income, occupation and housing type— is what is needed.
Changes in Bureau of Statistics procedures
What caused me to write this article was an email from an overseas friend who has been studying Fiji for several years. He wrote:
“ I've nearly completed data collection for the book … and am currently updating and revising the Fiji article … In the process of revising my Fiji stats, I became very concerned by apparent changes in FIBS (Bureau of Statistics) data collection and compilation of HIES (Household Income and Expenditure) reports that present data that are not comparable from report to report.”He had several concerns — no thorough analysis of census data since 1996; statistics based on “extrapolations of extrapolations;” Overseas Merchandise and Trade Statistics that are now inconsistently tabulated and reported — and so on. But his main concern was the HIES reports that readers will recall are used to analyse poverty and make recommendations on minimum wages. This is what he writes:
“I am most alarmed by the glaring omissions and inconsistencies in the HIES Reports. Data that should be presented consistently and comparably over time are either omitted or smoothed beyond appropriate use. (Data on Indian emigration that suppresses their earnings and inflates Fijian earnings is not noted or presented.) There has been no consistent presentation of income deciles by ethnicity, weekly household earnings by ethnicity or other indicators of social well being since the 1997 UNDP. (We really do not want to present data that shows growth in the Fijian and Indian upper income deciles.)He concludes with a warning against the use of statistics to act out researchers’ biases, and thinks “maybe the Fiji government should just save money and refrain from paying for surveys until they have sufficient funding to do it right.”
“And the latest HIES Poverty Report is a joke; far from a meaningful "gaps" report that adequately measures dispersion of household incomes and expenditures or poverty between and within groups by household composition, size or ethnicity.”
He is not saying that people are deliberately misusing statistics; merely that too many are incomplete, questionable, and cannot be compared with earlier periods. Removing —or not using— race as a major explanatory variable is a major part of this problem.
A knowledge-based society
Pillar 9 of the People’s Charter calls for Fiji to become a knowledge-based society. Raw data and statistical analyses form part of the bedrock from which this and other Charter goals can be planned, monitored and achieved. Race/Ethnicity needs to be put back into the statistics Government gathers because it is a Fiji reality , and refining race into sub-groups could make it even more useful. These also are part of Fiji’s multi-racial, multi-ethnic and multi-cultural reality.
* Race is biological whereas ethnicity is race, culture and acceptance by a respective ethnic group. Racism is the use of race to establish and justify a social hierarchy and system of power that privileges, preferences or advances certain individuals or groups of people usually at the expense of others. Racism is perpetuated through both interpersonal and institutional practices.