Skip to content

FIJI’S ELECTORAL SYSTEM AND ELECTION RESULTS Implications of Long-Term Population Change “Racial split causes worry: implications of population changes”. The Fiji Times, 27 April 1994. “Dead votes, ghost votes, and fair regional distribution”. The Fiji Times, 28 April 1994 “Racial or multiracial: where is the future?”. The Fiji Times, 29 April 1994.]

29/03/2012

“Racial split causes worry: implications of population changes”.

The Fiji Times, 27 April 1994.

 

  Introduction

Public discussions of the Fiji Constitution invariably focus on the detailed operation of the Electoral System, and in particular, the patterns of distribution of the number of Parliamentary seats.

While there has been some discussion of the distribution between different groups of Fijians, or parts ofFiji, and of possible alternative voting systems (cross-voting, common roll, and proportional representation), the major bone of contention has been the split in the number of parliamentary seats, between the different ethnic groups inFiji: indigenous Fijians, Indo-Fijians and Others.

Underlying most of the inter-ethnic political tensions, always tacitly understood but often explicitly expressed over the last few decades, has been the fear that indigenous Fijians would be numerically swamped by Indo-Fijians if Fiji were to have any form of proportional voting electoral system, including at the extreme, a common roll, or one-person: one:vote system.

With one central function of political power being the allocation of public resources, there has also been the associated tacit fear, that once in political power, Indo-Fijians would allocate to themselves, more than a “fair share” of public resources.

Indeed, Fijians in political power for the last two and half decades, have attempted to redress ethnic imbalances in economic involvement through ethnically discriminatory policies in education and public sector employments opportunities, which themsevels have created deep dis-satisfaction amongst Indo-Fijians in the country.

However, how real are ethnic Fijian fears, of numerical political domination by Indo-Fijians?

This paper argues that the demographic trends over the last decade in Fiji, strongly indicate that Fijians are not likely to be swamped by Indo-Fijians in the foreseeable future, either in numbers of voters, numbers of children vying for primary, secondary and tertiary education, or numbers of school leavers applying for jobs.

The reverse appears to be the case, with indigenous Fijians expected to dominate at all age levels within ten years.  The demographic changes taking place, in fact, will fundamentally alter the nature of both the political and economic tensions between various groups in the very near future.

It is important that the public have access to hard objective data, to ensure that there is little ground for racist demagogues to inflame ethnic fears and push ordinary persons into extreme political positions, which leave no room for compromises, so essential for any consensus on a nationally acceptable constitution.

The demographic changes taking place over the next ten years, are presenting those in power with a golden opportunity to establish policies, based on sound unbiased principles, in crucial areas such as scholarships, public sector employment, and inFiji’s electoral system.

Ethnically unbiased policies on the Fiji Government’s part, will not only go a long way to heal the current ethnic divisions in our society, but in terms of the electoral system, they are the likely minimum requirements for re-entry ofFijiinto the Commonwealth.

The demographic changes taking place inFiji, are now presenting Fijian power-brokers with a rare opportunity to rebuild the trust of a nation.

 

B       THE 1994 ELECTIONS: some inequities

 

Inequality of Parliamentarians: Effective Voters per Parliamentarian

In the Constitution, and in the day to day decision-making inFiji’s Parliament (leaving aside the constraints on non-Fijian parliamentarians who for example may not become Prime Minister), each parliamentarian is considered the equivalent of every other one.

When the President has to appoint the Prime Minister, and has to decide who has the majority support of the Members of Parliament, this is done by simply counting the numbers of parliamentarians supporting the competing candidates.  A parliamentarian from a large constituency does not count for more than the parliamentarian from a smaller constituency.

It is also presumably the case that each Parliamentarian or Cabinet Member has one vote in decision-making.  Each parliamentarian presumably also counts as one voice in decisions regarding the allocation of resources, including those affecting his or her constituency.   The strong implication is given that if the majority of parliamentarians vote in one direction, they have associated with them the majority of the voters of the country.

Yet, how many voters do these parliamentarians represent?  How “equal” are these parliamentarians?  Or is it, as in Animal Farm, that some parliamentarians and constituencies (and even losing candidates) are “more equal” than others?

It is important to note that it is not enough to simply compare the “gross votes” cast for candidates: some Fijian constituencies had multiple seats with each voter being able to vote for multiple candidates.  The number which must therefore be calculate for purposes of comparison in Fijian seats, “Voters per Seat” (ie the Total Registered Voters divided by the number of parliamentarians being elected from that constituency).

It therefore follows that if it takes 2000 persons to elect a parliamentarian from constituency A, and 10,000 from constituency B, then for constituency B, 8000 voters may as well not exist, they may as well be “dead”.

What should be a reference seat size?   We cannot go by the smallest constituency (e.g. the tiny General Constituencies) or we would end up with a massive parliament.

One logical point of reference, which we can use to analyze the constituency registrations, is the Fiji-wide average number of registered voters per parliamentarian, which was 4716 in 1994.

Those seats with above-average (compared to 4716) numbers of voters will effectively have “dead” voters (representing the excess), while others will have “ghost” voters where the actual number of voters is less than the national average.

We will not discuss further the ghost voters in the General Constituencies, which all have so many “ghost” voters that they could be referred to as the “Twilight Zone”.  (Their current preferential situation is a relic of the colonial past when the economically and financially influential whites gave themselves a balance of power out of all proportion to their actual numbers).


“Dead votes, ghost votes, and fair regional distribution”.

The Fiji Times, 28 April 1994

  

Dead and Ghost Voters

It is abundantly clear that many constituencies have an unfairly large number of voters, effectively wasting or neutralizing large numbers of voters.

Thus the Fijian Urban Tailevu/Naitasiri constituency (with 16,221 voters per seat) had 11,500 extra, wasted, or “dead” voters: this constituency was clearly entitled to another 2 parliamentarians.

Suva City Urban Fijians had 7,500 “dead” voters, entitled to almost another two parliamentarians.  Fijian Urban Western, the Indo-Fijian seats for Labasa Rural,

Macuata East and Ba Urban all had enough “dead” voters to warrant representation by another parliamentarian each.  This issue may be more systematically investigated as the range in percentage “deviations” in seat sizes.

Large Deviations in Seat Size in 1994

It is to be expected that there will be some variation in the number of voters per constituency: it would be impossible to have exactly the same number for each, if for not other reason than difficulties in having ideal physical boundaries.  But how much deviation may be expected on should be allowed?

For the 1987 election, the number of registered voters in each constituency for the Fijian seats, showed less than 34% deviation from the average, while for Indo-Fijians, it was less than 20% (Appendix A, Table 1).  The average constituencies for Fijians and Indo-Fijians was around fourteen thousands (Appendix A, Table 2).  Proportionately on an ethnic basis, Fijians and Indo-Fijians would have had 13 seats each (instead of 12) while the Generals would have had 1 only, instead of 3 under the older system.

For the 1994 election, the deviations from the national average were massive, for both Fijians and Indo-Fijians.  The percentage deviation for Fijian seats went from minus 77% for the small seats, to plus 244% for the larger seats (Table 3, Appendix A).  That for Indo-Fijian seats went from minus 68% for the small seats to plus 117% for the larger seats.

Individual Injustices

The gross discrepancies in the sizes of the constituencies and the weaknesses in the electoral sytem, have had the inevitable result that quite a few politicians, despite massive electoral support, failed to win a seat while others with considerably fewer votes, “got in”.

It must be noted that the gross votes (as recorded in most media reports) are not the appropriate yardstick in Fiji’s electoral system.  The measure to compare each candidate with is “net votes”, which equals the number of gross votes received, divided by the number of seats in that constituency.

Discrepancies will inevitably be there for Indo-Fijian seats, because the bias against them in the national aggregate allocation, must inevitably inflate some of their constituencies.  Yet such discrepancies are unusually large in the Fijian constituencies also.

Such discrepancies can lead to inconsistent positions in politics. Thus there was some opposition within Government, for a Cabinet position for an experienced politician who lost in Lau, supposedly because he had “failed” to be elected.  Yet he obtained more votes (net) than at least of 5 other candidates who “won” and got into Parliament.

This inconsistency in the system is clearer in the Fijian Tailevu/Naitasiri constituency where an experienced losing candidate obtained 2,509 net votes, which was higher than that received by 48 “winners” in the 1994 Parliament.  The losing STV candidate for Nadroga had 1651 net votes, which was more than that received by 20 parliamentarians successful in 1994.  The three losing ANC candidates for Ba had around 1500 net votes, again more than that received by 18 current parliamentarians.

When the votes of such losing candidates are aggregated for parties, it is quite possible for parties with significant national support to not receive their fair share of the seats in Parliament.  Thus both the ANC and the Nationalists should have received at least 2 additional seats in Parliament.  Some element of proportional voting is clearly needed to rectify this.

While some of the discrepancies may be created by vote-splitting if a large number of candidates stand for election in any one constituency, the biases are much deeper, and caused ultimately by the biases in the constituency boundaries.

These biases may be seen if we group the votes according to regions, or islands, or types of constituency, work out the numbers of “dead” and “ghost” voters, and look at the transfer of seats required from those categories with too many “ghost” voters to those with “dead” voters.

Ethnic Redistribution

In a previous section, we have discussed the likely redistribution required for the 1994 election, if decided by the over 20, with indigenous Fijians probably reducing their number of parliamentarians by 3, and the others by 2.  Indo-Fijians would therefore have gained an additional 5 parliamentarians.  This is probably the most accurate redistribution among the ethnic groups.

It is unfortunate, however, that prominent parliamentarians and political leaders in the Opposition seem to think that the only categories discriminated against in the Electoral system are Indo-Fijians.

What the electoral data makes clear is that there are other significant parameters requiring a redistribution of seats, both within Fijian and Indo-Fijian constituencies.

Because the analysis that follows from here on is by “registered voters” and not by potential voters, there is a small downward bias for areas with relatively higher indigenous Fijian and Other populations (who seem to under-register for elections).

Rural and Urban Redistribution

One clear bias in the electoral system is the redistribution between the rural and urban peoples.

Urban Fijians, given their numbers of registered voters, ought to be electing an additional 5 parliamentarians.  Given that the urban Fijians are far more likely to be involved in the cash economy and to be paying taxes, it would seem to be economically unjust to deny them their proper representation.

Table 4   Required Changes in Seats for 1994 (Rural/Urban)

———————————————————————————————————

Ghost                 Dead               Required Change in Seats 

Votes               Votes               Unadjust.         Adj. for Under-reg

———————————————————————————————————-

Fijian Prov.         40374                                                  -9                     (-8)

Fijian Urban                                -21556                         +5                   (+5)

Indian Rural                              -13290                         +3                   (+2)

Indian Urban                               -18858                         +4                    (+3)

Others             13365                                                  -3                     (-2)

———————————————————————————————————-

Of those required for Indo-Fijians, 3 would be for urban areas, and only 2 for rural areas.

By Island Groups

Aggregating roughly by the main island groupings (which is difficult for some constituencies because of the way in which they are constructed with bits and pieces from all over the place), again reveals the under-representation ofViti Levuvoters.

Table 5   Required Changes in Seats for 1994 (by Islands)

————————————————————-

Ghost         Dead            Req. Change

Island Gr.    Voters         Voters            in Seats

————————————————————-

Small Islands      14923                                      – 3

Vanua Levu         4095                                      – 1

Rotuma                  1028

Viti Levu                                            -20011                  + 4

————————————————————

It should be noted thatViti Levuis the locus of the major industries (sugar, tourism, gold and agriculture), as well as most of the commerce and government.  It is unusual that it is deficient in parliament by a full 4 seats.  WithinViti Levu, there is also a clear regional bias.

The next table makes evident that extra seats required for Viti Levu, would need to go to the West, where there is the largest numbers of currently “dead” voters, and where the sugar, tourism and gold industries, the lifelines ofFiji’s economy, are concentrated.

Table 6   Required Changes in Seats for 1994 (by region)

————————————————————————–

Ghost                 Dead            Change in

Region        Voters               Voters             Seats

————————————————————————-

Central                                   ‑214

East            16641                                                           – 4

West                                   ‑17420                      + 4

Rotuma        1028

——————————————————————–

It must be noted that these discrepancies outlined above, will not disappear with proportional representation, unless either the regulations or the relevant party rules ensure that the unelected candidates with the largest number of votes would be chosen for parliament as part of the proportional quota.

With a bit of effort, it is clearly possible to draw up any number of such cross tabulations, with corresponding demands for balance.

However, these comparisons would be completely unnecessary if the Electoral Boundaries Commission simply ensured that each constituency had roughly the same number of eligible voters per seat.

The numbers of voters in 1999 are likely to be totally different from that obtained through the 1986 Census, given the significant demographic changes which are taking place throughoutFiji.

It is extremely important, therefore, that the Bureau of Statistics, obtain accurate projections of the likely population distribution throughoutFiji, in order to assist in accurate and sensible electoral boundary determination.

This should be possible by using the results of the 1996 Census, and getting professional inputs by demographers to make reliable projections.

Constituency or the Nation?

Much more difficult to address are the peculiar problems of accountability created by our electoral system.  Who does the “successful” candidate represent?  Only those who voted for him or every one in his constituency?

If he makes it to the Cabinet, does he still represent only his constituency, or perhaps all the voters who voted for his Party?  Or does he represent all people, in all constituencies, regardless of the degree of their support for his Party?

In the existing electoral system, parliamentarians can only be elected by voters in particular constituencies.  These voters justifiably feel that “their” parliamentarian should fight for “their” interests, or at least of their constituency.

This pressure is there even if they become part of the Cabinet, with ministerial responsibility for entire sets of national policies, such as in education, health etc.

If the Minister ignores the interests of his voters in the constituency, he will face difficulties in getting elected the next time around.  On the other hand, if the Minister does not adequately fulfil his national responsibilities, he is likely to lose the job in the next Cabinet reshuffle.

This contradiction may be lessened if there are some Parliamentary seats elected by proportional representation, with Parliamentary positions thus depending on all the votes received by a party throughout the country.

What of the Losers?

InFiji’s polarized politics, it seems to have been taken for granted that the Party in power, first and foremost, looks after the interests of those that voted them in.  The interests of the voters who supported the parliamentarians not forming the government are easily ignored.  This has been the situation for most Indo-Fijians, but only from the mid-1970’s when public policies became ethnically polarized.

The current electoral system fosters this position, since all that matters is for the government of the day to have the support of a simple majority of the votes in the House of Representatives.

However, it is clear from the data, that the Government of the day does not necessarily even have the majority of the voters behind it.  Thus the SVT only had 30 percent of all the 1994 voters, and around 35 percent, counting its coalition partners, the GVP and all the Independents.

Yet the Government of the day, allocates revenues derived from all the tax payers in the country, towards its own priorities in infrastructure development, social welfare, health, and education.  These priorities are now generally those of the Government’s supporters.

However, should the other (and indeed the largest) contributors of the tax revenue, not have any say at all in its expenditure?  Should those in Government not at all be accountable, through the ballot box, to the largest generators of the tax revenue?

Unfortunately, these results are the likely consequences of the racially compartmentalized electoral system where each voter is only allowed to vote for candidates of his or her alleged “ethnicity”.

Although it must also be noted that they are not the only consequences.  There is nothing in the 1990 Constitution to stop any political party from putting up a candidate in any ethnic compartment.  That they are not doing so, is a consequence of their narrow racial polarization.

Racial Political Parties

It is an extremely unfortunate consequence ofFiji’s political history, only partly contributed by the 1990 Constitution, that the existing major political parties, with the exception of the ANC, have become almost completely racially polarized.  The situation is worse now than it was in the early 1970s, or even at the 1987 elections.  All major political parties are responsible for their own narrow racialism.

From their inception, and by explicit design, the ruling SVT Party, the Nationalists, the STV, and the Fijian Association Party, are all openly of indigenous Fijians, for Fijians.  The GVP is for the “Others”.

The NFP, made some feeble attempts in the early seventies to woo some Fijian votes.  They did obtain some impecunious prominent Fijians who boldly stood with them, and did manage to obtain a few percentage points of Fijian votes now and then.  A few radical political Fijians (still prominent in politics) also joined them, but were discarded or driven out by the narrow visions of the leadership.  The NFP is now unabashedly an “Indian” party, not even making any pretence of interest in Fijian voters.

The Labour Party, at its inception, and going by its rhetoric, was seen by its supporters as the only genuinely multiracial party.  Quite revealing, however, was that its several credible Fijian leaders, chose to stand in Indo-Fijian dominated National seats, where they won with very little Fijian votes.  Nevertheless, the Labour Party’s few communal Fijian candidates, with very little financial support from the Party, managed to win a significant 9% of the indigenous Fijian votes in 1987.

However, the Labour Party leaders, together with the overt or tacit support of their intellectual advisers, all driven by their zealous and petty power politics, gradually marginalized potential indigenous Fijian Labour Party leaders and Fijian voters.

While the coups made abundantly clear that no winning party could remain as the Government unless it had a significant proportion of Fijian votes, the Labour Party leaders have proceeded to reduce it to, and indeed openly refer to it as, an “Indian” Party, representing Indian interests.

Now the leaders of both the NFP and Labour Party persistently argue that the only people significantly discriminated by the 1990 Constitution are people of Indian origin.  They largely ignore the marginalization of the urban Fijians and theWestern Fijians.  They also ignore that economically and materially, in terms of the standards of living, the bulk of the indigenous Fijians are as, if not more poverty stricken, than the worst of the Indo-Fijians.

Whenever a progressive posture is called for (especially when rousing international support) the rhetoric of multiracialism is trotted out.  Yet in the 1994 Election, very few NFP or Labour candidates contested Fijian or General seats, and little financial support was given to these candidates by these Parties.  Together, they won less than half a percent of the Fijian votes.

Both the major opposition parties lack the vision, the capacity, or the will to front up to all the ethnic groups in the country, for electoral support.  They are incapable of being a Government, and many of their leaders (except for those without alternative jobs) are probably quite grateful for that.

In hindsight, the now defunct Alliance Party, has been the most successful multiracial party.  Always with more than two thirds of Fijian support, it also won 24 percent of the Indo-Fijian votes in 1972, and while the support declined over the years, their Indian support did not drop below 14 percent.  The Alliance Cabinet also had Indo-Fijians in Cabinet, rising from around 3 in the early 1970s to 5 in 1978 and then gradually declining to 2 by the middle eighties.

None of the existing major parties have demonstrated any capacity to be a genuinely multiracial national party addressing the whole nation’s needs.

Gains from Multiracialism

Ethnically polarized parties must come to terms with the fact that by limiting themselves to one ethnic group, they are always going to place rigid upper limits to the amount of total voter support they can muster.

The Fijian Association Party, which won a significant 15 percent of the Fijian votes, will always be limited in its national support, as long as it remains a “Fijian” Association.

In the 1994 Election, the SVT had the support of only 30% ofFiji’s voters.  One important reason for this dismal performance was that it contested only Fijian seats, and obtained only 65% of the Fijian votes.

The Alliancedropped to this low figure once in their history, in the first 1977 election.  It should be remembered that in their losing 1987 election, the 15% of Indo-Fijian votes it won, enabled it to have the support of a higher proportion ofFiji’s voters, than did the winning Coalition.

It should be remembered that the poorer classes among the Indo-Fijian moved away from theAlliance(after their significant 25% support in 1972) when they began to be denied scholarships, and were discriminated against for entry to, and promotion within, the Civil Service.

Yet these are issues on which policies based on equity principles, are guaranteed to win any Party support from the masses of the Indo-Fijian population.

Regardless of the grand ethnic rhetoric of Opposition Parties, there is no pure “ethnic” division in the country.  There are well-off Indo-Fijians (businessmen and others) who are doing very well under the SVT, and contribute generously to the key power-brokers.  Their children have no difficulty getting scholarships or jobs or other opportunities despite the 1990 Constitution.

The really discriminated people are the poorer classes of all races, whose major concern is not pieces of paper certifying their equality (in poverty) before the world, but bread and butter issues such as education for their children, and jobs and incomes.

Any political party which addresses these problems on the basis of equity, will attract their votes, including those of the Indo-Fijians, as did the Alliance Party in the early seventies.

Failure of the Opposition Parties and Communities

It is to be regretted, but is very revealing, that the numerous non-Fijian Education societies, and the Opposition Parties, despite the presence of a large number of extremely wealthy supporters, have yet to establish a National Scholarships Fund for needy Fiji citizens who are classified as non-indigenous, although some no doubt help on an individual basis.

Yet Indo-Fijian political donations, to all parties, seem to be made with great ease, as in the 1992 and 1994 elections.  Hundreds of thousands of dollars are spent with great alacrity in fighting largely irrelevant battles for the sterile leadership of the marginalized Indo-Fijian community.

It must be noted that well-off Indo-Fijians currently, and especially over the next ten years, will have even greater amounts of disposable income than indigenous Fijians.  While Fijians currently have 63% of their population as dependents (under 15 or over 65), Indo-Fijians have only 52%.  By 2004, the proportions will have declined to 56% for Fijians and 37% for Indo-Fijians.

These differences in disposable income do not seem to be matched by generosity in scholarships at the national level.  While some economists may not approve of money being collected from the poor for distribution by the well-off, the relatively poorer Fijians accumulate hundreds of thousands of dollars in solis.

Indo-Fijians political leaders, despite the massive problems in education faced by their poorer sections, have yet to demonstrate any such collective concern, except for futilely getting more of their candidates into Parliament.

It seems that the only recourse for poorer Indo-Fijians and others, is the Government of the day.  It is therefore fortunate, but critically important for the public, to understand that the demographic changes described above, will facilitate the efforts of any Government to put in place such equitable policies, to ensure that the poorer member of all ethnic groups have access to scholarships and public sector employment.

 

B:        IMPLICATIONS OF POPULATION CHANGES

There are two existing sets of population projections, one by Seniloli (1993) and the other by the Fiji Bureau of Statistics (1989), both using the 1986 population as the projection base.

As far as Fiji’s total population is concerned, both these sets of projections are reasonably accurate, if judged by how close they are to the most recent population estimates prepared by the Bureau.

However, the projections differ significantly from recent estimates for the different ethnic components.  The major causes of the differences are that the assumptions underestimated the likely continuing rates of emigration of Indo-Fijians (hence over-estimated the Indo-Fijian population), and probably also underestimated the fertility rates of Fijians (hence under-estimated the indigenous Fijian population).

For these articles, the population projections were derived by using the Bureau’s 1989 estimated population breakdown as the base, and using the PEOPLE software to project into the future.

The demographic parameters used were the medium variant parameters used by Seniloli, with the exception that the overall level of migration was changed to more closely resemble the net departures over the last few years.

Key Demographic Trends

It is well known thatFiji’s population growth rate has declined from an annual rate of 3.2% between 1956 and 1966 to 2.0% between 1976 and 1986.

Table 1           Percentage of Fiji’s Total Population

——————————————————————————–

1986    1989    1994    1999    2004    2009

——————————————————————————–

% FIJIAN                 46          49        51        53        55        57

% INDO-FIJIAN    48          47        44        42        40        38

% OTHERS             6             5          5          5          5          6

TOTAL                  100     100     100     100     100     100

——————————————————————————–

TOT POP (000)       715     726     777     828     878     926

——————————————————————————–

However, the next decade will probably see the annual growth rate decline even further to around 1% per annum, largely because of continued out-migration, and low and declining fertility rates of Indo-Fijians.

While the growth rate for indigenous Fijians is expected to decline from around the current 2.1% to 1.8% between 2004 and 2009, the growth rate for Indo-Fijians will decline from around 0.2% currently, to possibly -0.1% over 2004-2009.

This means that while indigenous Fijians were 46% of the population in 1986 (as compared to 49% for Indo-Fijians), their share will rise to 53% by 1999 (Indo-Fijians 42%) and 57% by 2009 (38% for Indo-Fijians).

In any electoral system, however, it is not the population which votes, but those 21 and over.

The Voting Group (Over 20)

Because of the peculiar demographic patterns of the last few decades, the ethnic proportions of the Over 20 Age Group, has been somewhat different from the proportions of the total population.

This would have been indicated by the relative voter registrations for this year’s elections, where Indo-Fijians still out-numbered the indigenous Fijians.  But even the voter registrations would have been misleading.

The following Table (with the expected percentage breakdown of the voting-age population over the next fifteen years) indicates that even by 1992, the number of voting age indigenous Fijians should have outnumbered the number of voting age Indo-Fijians.  By 1994, indigenous Fijians comprised 49 percent of the Over 20 age group (even though they were only 47 percent of the registrations).

In terms of numbers, in 1992, there were probably some 5000 more indigenous Fijian potential voters than Indo-Fijian, with the difference rising to about 13,000 in 1994, and probably 26,000 by the next election in 1999.

Table 2   Proj. Perc. of the Over 20 Age Group (%)

———————————————————————

1991    1994    1999    2004    2009

———————————————————————

Fijians                                 48          49        50        52        53

Indo-Fijians             47          46        45        43        41

Others                       5           5          5          5          5

TOTAL                   100      100     100     100     100

———————————————————————

Since the actual number of registered voters was higher for Indo-Fijians than Fijians, it would seem that a much larger proportion of indigenous Fijians (some 22%) do not register to vote compared to Indo-Fijians (15%), although about the same proportion of those registered actually do vote.   This under-registration is probably explained by the much more scattered and isolated nature of indigenous Fijian rural communities.

Overall equity, however, requires that calculations of the desirable seat representation in Parliament, should use the Over 20 Age population and not just the numbers of registered voters (as some political analysts seem to be doing).

Ideal Seat Distribution

If the number of seats in Parliament are distributed according to the proportion of the Over 20 population of the major ethnic categories,, and the size of the Parliament is maintained at 70, then the following numbers are projected.

Table 3       Ideal Number of Seats (in proportion to Over 20 Age Group)

————————————————————————

1989     1991     1994      1999     2004     2009

————————————————————————

Fijian                           33       34          34          35        36          37

Indo-Fijian       33       33          32          31        30          29

Others                4         4            4            4          4            4

TOTAL            70       70          70         70        70          70

————————————————————————

Thus by this year’s election, the indigenous Fijians would already have out-numbered Indo-Fijians by 2 seats, the difference rising to 4 by 1999, and to 8 by 2009.

It is clear therefore that a basic egalitarian electoral system based purely on the numbers of potential voters, which is probably the minimum that would be required before the re-admission ofFijito the Commonwealth, would already give indigenous-Fijians a numerical majority in Parliament.

That the majority is not large is actually immaterial.  The Machiavellian lessons from the 1987 coups, and the emerging demographic changes, mean that it is extremely unlikely for there ever to be a Government in power, which is not comprised of Fijians in the majority.

A major issue for any constitutional discussions ought to be provisions allowing for an equitable sharing in the exercise of Goverment power amongst the major interest groups.

These interest groupings are not just along narrow ethnic lines, but also by regions, and types of constituencies.

 

C         IMPLICATIONS FOR CALL ON PUBLIC RESOURCES

The demographic changes taking place are going to have a profound impact on the tensions in society, arising out of public demands for funds for education, scholarships for higher education, and access to public sector jobs.

Because of the emigration and low and declining fertility of Indo-Fijians, the total number of primary age students (age 6-11) in Fiji will probably decline by some 5,000 between 1994 and 1999.

This is almost unheard of in developing countries, although similar trends are also present inTongaandWestern Samoa, two otherPacificIslandcountries with high out-migration.

Given that the primary school budget is by far the largest component of the education budget forFiji, the above projection of total potential primary enrolment indicates that there will be some easing of pressure on the Government budget, due to the levelling off of the primary age population.

Table 7        Primary Age Population (000 and %)

——————————————————————–

AGE GROUP                   1994     1999     2004    2009

——————————————————————–

Fijians  6‑11                        57       60       64      68

Indo-Fijians 6‑11                47       39       36      34

____________________________________________

Total  6‑11                          104      100      100     102 *

_____________________________________________

% Excess of Fij. overInd.      21       54       81      98

——————————————————————–

[* Total ignores other ethnic groups who are assumed to grow at about 2% per year. Errors due to rounding.]

An important point to note is that the number of Indo-Fijian primary age population will show a massive decline in absolute terms by almost 30%, so that indigenous Fijians will outnumber Indo-Fijians by 54% in 1999, and 98% in 2009.

The reduction in Indo-Fijian enrolment will also create an enormous capacity in what are currently Indo-Fijian primary schools, for enrollment of children of other ethnic groups (and associated benefits), for improvement of pupil:teacher ratios and the quality of teaching generally.

Similar changes are also likely to take place at the other age groups, although the timing may be slightly different.  Total populations will be declining in the 12-15 age group (junior secondary) from 1994, for 16-17 (senior secondary) from around 1999, and for 18-20 from 2001.

Of interest again is that Indo-Fijian population at the secondary school level (corresponding to Form 1 to Form 6) will decline in absolute terms by 31% between 1994 and 2009, while the corresponding Fijian cohort will grow by 21%.  The effect on total population in the 12-17 group will be a decline from 1999 onwards.

    Table 8  Projections of School Age Population (000)

————————————————————————–

1994      1999       2004       2009

————————————————————————–

Fijians 12-17 (000)             52            56         59             63

Indo-Fijians 12-17 (000)    49            46         38             34

Total 12-17 (000)               101        102         97             97

________________________________________________

% Excess of Fij. overInd.      6           22          55             85

_________________________________________________

These trends all indicate a future easing of pressure for resource requirements of education at the secondary levels.

They also indicate the opportunity for those in the Ministry of Education to experiment with the placing of Fijian students in good schools currently dominated by Indo-Fijian students, to see if there are flow-on effects on the academic performance of Fijian students.

Table  9   Projections of 18-20 Age Groups (000)

————————————————————————-

1994      1999       2004      2009

————————————————————————-

Fijians 18-20 (000)                 23        26            28          30

Indo-Fijians 18-20 (000)          21        23            21          17

Total 18-20 (000)                     44        49            49          47

________________________________________________

% Excess of Fij. overInd.  10        13         33        76

________________________________________________

Table 9 indicates the similar trend atthe 18-20 age group which may be taken as those cohorts which imply pressure on the job market.   Again, the Indo-Fijian cohort will stabilise after 1999 and eventually decline.  Relatively speaking, Fijians already outnumber Indo-Fijians by 10% in 1994, probably by 13% in 1999, and by 33% in 2004.

The projection data implies ever-widening gaps between Fijians and Indo-Fijians at all these cohort levels discussed above, representing reduced competition on resources for primary, secondary and tertiary education, and the job market.

On the basis of these population projections, the argument that Indo-Fijians will continue to dominate indigenous Fijians through sheer numbers of population, is just not tenable.

These projections also indicate that Government may adopt public policies in a number of areas, based on equity, rather than ethnicity, without necessarily compromising indigenous Fijian interests.

Scholarships on Principle

Over the last fifteen years, many social commentators have called on the Government to maintain ethnic parity in the awarding of scholarships.

While the above would surely be easier to do over the next ten years, what is far more important is that Government establish sound principles for the awarding of scholarships, not based on ethnicity, but on need, and regardless of the impact on ethnic balance.

Successive Governments, no doubt driven by the interests of their wealthy supporters (of all ethnic groups), have steadfastly refused to put in place a means test for scholarships.

Scholarships at all levels, starting from secondary education, should be on the basis of need, with an appropriately designed means test.  Currently, undeserving students of all ethnic groups receive scholarships, even where their parents are fully able to finance the education of their own children.

Secondly, the awarding of all scholarships (overseas or regionally) at the post-secondary level should be on a shared cost basis, with students repaying a certain proportion of the award back to Government following graduation and employment.

Thirdly, there should be a proper unified scholarships awarding Committee, which brings together all the separate bodies such as the Education Department, the PSC and the FAB, as well as representatives of the non-Government education bodies.

This would enable scholarships to be awarded in the areas of national need, and according to a sound set of consistent principles.  The current mechanisms leave a lot to be desired, both on the grounds of efficiency and equity.

Proportionate Job Opportunities

As important to the poorer people as the provision of scholarships, is access to employment opportunities, and equity in promotions: in the public sector, police and military.

Again, the demographic changes taking place indicate that there are no population grounds for the continuation of discrimination on the basis of ethnicity.

The authorities might do well to consider also giving preferential access to applicants with parents below certain levels of income and wealth, again regardless of the implications for ethnic balance.

Genuine Democracy

Last but not least, there seems to be a need for genuine democracy in all the political parties.  While some political parties mouth their concern to restore democracy inFiji, their internal power politics indicates that they are as ridden by dictatorship as any of the other parties they criticize.

Political leaders focus too easily on the more obvious result of genuine democracy, which is that there is greater accountability of leaders to their people.

Political leaders usually resist, having great contempt for the abilities of their ordinary rank and file supporters to “understand what is required”.

Yet in many developing (and developed) societies, it is the nature of human beings to want leaders who will make their difficult decisions for them.  They will idolize those who succeed, and crucify those who fail.

Fiji’s history has many examples of leaders who have been berated by their peoples for failure and sellouts, and leaders who have berated their people for their ingratitude: the original Cession of Fiji; the 1970 Constitution; ALTA; the 1990 Constitution, and no doubt more in the future.

Political leaders may ponder on one completely ignored consequence of genuine democracy which is that it forces people out of their political laziness, makes them take full part in decision-making, forces them to learn from their mistakes, and ultimately take full responsibility for their own future, and that of their children.  There are no more scapegoats.

One genuine beginning of faith in our people could be a full referendum for the next Constitution.

 

Bibliography

Seniloli, Kesaia (1993) “Fiji: Population Projections, 1986-2010”. Pacific 2010: Challenging the Future, National Centre for Development Studies,AustralianNationalUniversity.

Fiji Bureau of Statistics (1989) Report on Fiji Population Census 1986, Occasional Paper 1: Population Projections for years 1986 to 2011, Bureau of Statistics,Suva.

Fiji Bureau of Statistics (1993) Current Economic Statistics,Suva.


APPENDIX

Table *     Deviation of Communal Seat Size from Average 1987

_________________________________________

Const.  Regis   Dev.   Perc.

Voters          Dev.

_________________________________________

FIJIAN      LAU/ROT  9338  ‑4779    ‑34

FIJIAN      NADROGA 11972  ‑2145    ‑15

FIJIAN      TAILEVU 12922  ‑1195     ‑8

FIJIAN      LOMAIVI 13288   ‑829     ‑6

FIJIAN      RA/SAMA 13396   ‑721     ‑5

FIJIAN      BUA/MAC 13625   ‑492     ‑3

FIJIAN      CAKAUDR 14796    679      5

FIJIAN      NAITASI 15007    890      6

FIJIAN      BA/NADI 15071    954      7

FIJIAN      REWA/SE 15517   1400     10

FIJIAN      VUDA/YA 16216   2099     15

FIJIAN      KADAVU/ 18250   4133     29

GENERAL     NORTH/E  1751  ‑1476    ‑46

GENERAL     WESTERN  2239   ‑988    ‑31

GENERAL     SUVA/CE  5692   2465     76

INDIAN      NAUSORI 11581  ‑2970    ‑20

INDIAN      TAVUA/V 12083  ‑2468    ‑17

INDIAN      LABASA/ 14147   ‑404     ‑3

INDIAN      SUVA/RU 14299   ‑252     ‑2

INDIAN      SIGATOK 14521    ‑30      0

INDIAN      BA      14599     48      0

INDIAN      SAVUSAV 14696    145      1

INDIAN      NASINU/ 14889    338      2

INDIAN      LAUTOKA 15167    616      4

INDIAN     SUVACI 15287    736      5

INDIAN      BA/LAUT 16269   1718     12

INDIAN      NADI    17073   2522     17

_______________________________________

Table *    1987 Constituency Characteristics

________________________________________________________________

AVERAGE   Perc. Dev.     NO    No of    Perc  Prop.

CONST.      Range      SEATS  Voters            NO.

________________________________________________________________

FIJIAN       14117   ‑34% TO +29%    12    169398     48     13

INDIAN       14551   ‑20% TO +17%    12    174611     49     13

GENERAL       3227   ‑46% TO +76%     3      9682      3      1

ALL          13100                   27    353691    100     27

________________________________________________________________

 

Table *       Deviation of Communal Seat Size from Average 1994

________________________________________________________________

Type      Constituency  No     Reg.   Vot./  Ext.    %     Party

Seats   Vot.   Seat   Vot.   Dev.

________________________________________________________________

F‑PROV.   NAMOSI         2     2167   1084   3633    ‑77   SVT

F‑PROV.   NAMOSI         2     2167   1084   3633    ‑77   SVT

F‑PROV.   SERUA          2     2853   1427   3290    ‑70   SVT

F‑PROV.   SERUA          2     2853   1427   3290    ‑70   SVT

F‑PROV.   LAU            3     6712   2237   2479    ‑53   FA

F‑PROV.   LAU            3     6712   2237   2479    ‑53   FA

F‑PROV.   LAU            3     6712   2237   2479    ‑53   FA

F‑PROV.   REWA           2     4591   2296   2421    ‑51   SVT

F‑PROV.   REWA           2     4591   2296   2421    ‑51   SVT

F‑URBA.   NORTHEAST      1     2339   2339   2377    ‑50   SVT

F‑PROV.   KADAVU         2     4981   2491   2226    ‑47   SVT

F‑PROV.   KADAVU         2     4981   2491   2226    ‑47   SVT

F‑PROV.   BUA            2     5193   2597   2120    ‑45   SVT

F‑PROV.   BUA            2     5193   2597   2120    ‑45   SVT

F‑PROV.   LOMAIVATI      2     6399   3200   1517    ‑32   SVT

F‑PROV.   LOMAIVATI      2     6399   3200   1517    ‑32   SVT

F‑PROV.   RA             2     7205   3603   1114    ‑24   SVT

F‑PROV.   RA             2     7205   3603   1114    ‑24  IND

F‑PROV.   MACUATA        2     7317   3659   1058    ‑22   SVT

F‑PROV.   MACUATA        2     7317   3659   1058    ‑22   SVT

F‑PROV.   TAILEVU        3    12855   4285    431     ‑9   SVT

F‑PROV.   TAILEVU        3    12855   4285    431     ‑9   SVT

F‑PROV.   TAILEVU        3    12855   4285    431     ‑9   SVT

F‑PROV.   CAKAUDROVE     3    13250   4417    299     ‑6   SVT

F‑PROV.   CAKAUDROVE     3    13250   4417    299     ‑6   SVT

F‑PROV.   CAKAUDROVE     3    13250   4417    299     ‑6   SVT

F‑PROV.   NAITASIRI      2     9265   4633     84     ‑2   FA

F‑PROV.   NAITASIRI      2     9265   4633     84     ‑2   FA

F‑URBA.   SERUA/REWA     1     5126   5126   ‑410      9   SVT

F‑PROV.   BA             3    16142   5381   ‑665     14   SVT

F‑PROV.   BA             3    16142   5381   ‑665     14   SVT

F‑PROV.   BA             3    16142   5381   ‑665     14   SVT

F‑PROV.   NADROGA        2    11615   5808  ‑1092     23   SVT

F‑PROV.   NADROGA        2    11615   5808  ‑1092     23   SVT

F‑URBA.   WESTERN        1     9179   9179  ‑4463     95   SVT

F‑URBA.  SUVACITY      1    12271  12271  ‑7555    160   SVT

F‑URBA.   TAILEVU/NAIT   1    16221  16221 ‑11505    244   SVT

G‑RURA.   RA ETC         1      815    815   3901    ‑83   GVP

G‑RURA.   BA/NADROGA     1     1852   1852   2864    ‑61   ANC

G‑URBA.  SUVACITY      1     2386   2386   2330    ‑49   GVP

G‑RURA.   SERUA          1     2964   2964   1752    ‑37   GVP

G‑RURA.   NORTHERN       1     3226   3226   1490    ‑32   GVP

(continued)

I‑RURA.   LAMI/KADAVU    1     1498   1498   3218    ‑68   NFP

I‑RURA.   BUA            1     1706   1706   3010    ‑64   NFP

I‑RURA.   CAKAUDROVE     1     2123   2123   2593    ‑55   NFP

I‑RURA.   MAGODRO/BA     1     4110   4110    606    ‑13   LAB

I‑RURA.   NADROGA E      1     4186   4186    530    ‑11   LAB

I‑RURA.   BA WEST        1     4194   4194    522    ‑11   LAB

I‑URBA.  SUVACENTRAL   1     4610   4610    106     ‑2   NFP

I‑RURA.   NAVOSA/REWA    1     4858   4858   ‑142      3   NFP

I‑RURA.   TAILEVU E      1     5023   5023   ‑307      7   NFP

I‑RURA.   CUVU           1     5562   5562   ‑846     18   NFP

I‑URBA.   NASINU E       1     5628   5628   ‑912     19   NFP

I‑URBA.   NADI URBAN     1     5749   5749  ‑1033     22   NFP

I‑RURA.   LAUTOKA S      1     5830   5830  ‑1114     24   LAB

I‑RURA.   TAVUA/RA       1     5863   5863  ‑1147     24   LAB

I‑RURA.   NADI RURAL     1     5959   5959  ‑1243     26   NFP

I‑RURA.   RA CENTRAL     1     6190   6190  ‑1474     31   NFP

I‑RURA.   LAUTOKA        1     6232   6232  ‑1516     32   NFP

I‑RURA.   NAWAKA/SABET   1     6352   6352  ‑1636     35   LAB

I‑RURA.   BA EAST        1     6432   6432  ‑1716     36   LAB

I‑RURA.   MACUATA W      1     7431   7431  ‑2715     58   NFP

I‑URBA.  SUVASUBURBA   1     7538   7538  ‑2822     60   NFP

I‑URBA.   NASINU N       1     7723   7723  ‑3007     64   NFP

I‑URBA.  LAUTOKACITY   1     7885   7885  ‑3169     67   NFP

I‑URBA.   NASINU‑S       1     8202   8202  ‑3486     74   NFP

I‑RURA.   MACUATA E      1     9126   9126  ‑4410     94   LAB

I‑URBA.   BA URBAN       1     9251   9251  ‑4535     96   NFP

I‑RURA.   LABASA         1    10219  10219  ‑5503    117   NFP

R‑RURA.   ROTUMA         1     3688   3688   1028    ‑22  IND

______________________________________________________________

Note: National Average Voters Per Seat = 4716.  For Fijians, the deviation around their own mean is even higher.

 

Table *      1987 Communal Results

____________________________________________________________

CONST                 TOT   ALL.   COAL.  OTH.   INF.  TOTAL

REG

____________________________________________________________

TOTAL FIJIAN       169398  92075  11217  14078   1130 118500

TOTAL INDIAN       174611  18798 100143   1670    559 121170

TOTAL GENERAL        9682   5106   1031    187     45   6369

TOTAL ALL          353691 115979 112391  15935   1734 246039

____________________________________________________________

Table *   1987 Communal Results (% of Communal Votes Cast)

_____________________________________________________________

RESP.

ALLIAN. COALIT  OTHER INFORM  VOTES  RATE

_____________________________________________________________

FIJIAN               77.7    9.5   11.9    1.0  100.0   70.0

INDIAN               15.5   82.6    1.4    0.5  100.0   69.4

GENERAL              80.2   16.2    2.9    0.7  100.0   65.8

TOTAL ALL            47.1   45.7    6.5    0.7  100.0   69.6

_____________________________________________________________

i.e.Allianceactually got more of the total votes than

did the Coalition, if voting was done communally.

TheAlliancewas more of a multiracial party than the Coalition

    Table *  1992 Elections: Response Rates

__________________________________

REGV/  NETV

SEAT   CAST

__________________________________

FIJIAN        142507 119996   84.2

INDIAN        147770 113721   77.0

GENERAL        12037   8008   66.5

ROTUMA          3571   2750   77.0

TOTAL         305884 244475   79.9

__________________________________

F‑PR  118363         87385   73.8

F‑UR  24143.         32611  135.1

I‑RU  101330         82013   80.9

I‑UR   46440         31708   68.3

__________________________________

NOTE: MY CLASSIFICATION OF INDIAN RURAL AND URBAN HAS PROBABLY CHANGED BETWEEN THE TWO ELECTIONS; BUT PATTERN APPEARS TO BE THE SAME

Table *  1994 Elections:  Response Rates

________________________________

REG      NET  % RESP

VOTERS  VOTES

________________________________

FIJIAN    155681 111589   71.7

INDIANS   159480 116950   73.3

GENERAL    11243   7977   71.0

ROTUMA      3688   2399   65.0

330092 238915   72.4

_______________________________

F‑PROV 110545  82402   74.5

F‑URBAN 45136  29188   64.7

I‑RURAL102894  72342   76.6

I‑URBAN 56586  44608   67.3

_______________________________

Note:     GULABDAS AND DIMURI were GIVEN VOTES IN PROPORTION TO THE RESPONSE RATES  OF INDIANS (73.3%) AND FIJIANS (71.7%) ALTHOUGH RURAL/URBAN DIFFERENCES QUITE ACUTE.

    Table *  1992 ELECTIONS: PARTY SUPPORT (NET VOTES)

_________________________________________

PARTY  FIJIAN INDIAN   GEN.   ROT.  TOTAL

_________________________________________

ANC      9762    500   2869         13131

FIC             1783                 1783

GVP                    5079          5079

IND      6952    387     60   2750  10149

LAB      1354  54292                55646

NAT     14353                       14353

NFP            56759                56759

STV      5165                        5165

SVT     82411                       82411

_________________________________________

TOTAL  119996 113721   8008   2750 244475

_________________________________________

PERCENTAGES

ANC       8.1    0.4   35.8           5.4

LAB       1.1   47.7    0.0          22.8

NFP             49.9                 23.2

FIC              1.6                  0.7

GVP                    63.4           2.1

NAT      12.0                         5.9

STV       4.3                         2.1

SVT      68.7                        33.7

IND       5.8    0.3    0.7  100.0    4.2

TOTAL   100.0  100.0  100.0  100.0  100.0

_________________________________________

Table *    ACTUAL AND PROPORTIONAL NUMBERS OF SEATS 1992

______________________________________

% OF     PROPO  ACTUAL  CHANGE

PARTY  VOTES      NO     NO     REQU.

______________________________________

ANC       5.4      4      0      4

LAB      22.8     16     13      3

FIC       0.7      1      0      1

NFP      23.2     16     13      3

GVP       2.1      1      5     ‑4

NAT       5.9      4      4      0

STV       2.1      1      2     ‑1

SVT      33.7     24     30     ‑6

IND       4.2      3      3      0

TOTAL   100.0     70     70      0

______________________________________

 

Table * 1994 Elections: Net Votes, Seats Won and Proportional

__________________________________________

PARTY   Net     Perc.   Prop. Actual Change

Votes             No    No     Req.

Rec.           Seats  Seats

__________________________________________

ANC     10945      5      3      1      2

FA      16936      7      5      5      0

GVP      4339      2      1      4     ‑3

INDEP    8174      3      2      2      0

LAB     51893     22     15      8      7

NAT      6613      3      2      0      2

NFP     64586     27     19     19      0

STV      3209      1      1      0      1

SVT     72222     30     21     31    ‑10

TOTAL  238915    100     70     70      0

Table *  1994: Parties’ Share of Votes

_____________________________

FIJ   IND    GEN

_____________________________

ANC         7     .4     36

LAB         0     44      2

NFP         0     55

GVP         0            59

SVT        65

FA         15

NAT         6

STV         3

IND         4      1      8

TOTAL     100    100    100

_____________________________

    Table *    Percentages of Communal Votes 1966-94

_________________________________________________________________

          Fijian Communal                Indian Communal 

Alliance    NFP/Lab      FNP      NFP/Lab     Alliance

__________________________________________________________________

1966       66                                  64          15

1968                                             77          21

1972         82          2                       73          24

1977 (1)     65                    25            73          16

1977 (2)     81                    12            85          14

1982         82          1          8            83          15

1987         77          9          5            82          15

  SVT                                           SVT

1992        69           1         12                       0

1994        65           0          6                       0

________________________________________________________________

    Table *   VOTER RESPONSE 1987-1994

_______________________________________________________________

> 20           REGISTERED VOTERS            % ch

@                                        1987-1994

AUG 86          1987     1992       1994

_______________________________________________________________

FIJIAN     159.9         169.4    142.5      155.7        ‑8

INDIAN     174.5         174.6    147.8      159.5        ‑9

OTHERS      19.3           9.7     15.5       11.2        15

TOTAL      353.5         353.7    305.9      330.1        ‑7

_______________________________________________________________

      VOTES CAST        

1987      1992      1994

_______________________________________________________________

FIJIAN                   118.5     120.0     111.6

INDIAN                   121.2     113.7     117.0

GENERAL                    6.4       8.0       8.0

ROTUMA                               2.8       2.4

TOTAL                    246.0    244.5      238.9

______________________________________________________________

        PERCENTAGES      

FIJIAN                     69.9     84.2      71.7

INDIAN                     69.4     76.9      73.3

GENERAL                    33.2     51.6      71.4

ROTUMA

TOTAL                      69.6     79.9      72.3

______________________________________________________________

COULD THERE BE MORE FIJIAN REGISTERED VOTERS IN 1987 THAN THE

NUMBER OVER 20 BECAUSE OF “OTHERS” SEEING THEMSELVES AS FIJIANS?

WHILE THEY WOULD  HAVE GONE TO THE OTHERS IN 1994?  NO: GENERAL DID

NOT RISE

ESTIMATES OF OVER 20 FOR INDO-FIJIAN GROUPS FOR 1992 AND 1994?

       Table *    Average Registered Voters Per Seat (by ethnicity and type)

______________________________________________________

FIJIAN  GENERAL  INDIAN  ROTUMA  ALL

______________________________________________________

F‑PROVINCIAL     3455                             3455

F‑URBAN          9027                             9027

I-RURAL                   2214    5415    3688    4810

I-URBAN                   2386    7073            6552

ALL              4208     2249    5907    3688    4716

______________________________________________________

Table * Average Registered Voters Per Seat (by ethnicity and island)

_________________________________________________________

FIJIAN   GENERAL  INDIAN   ROTUMA      ALL

_________________________________________________________

ISLANDS          2585                                2585

ROTUMA                                      3688     3688

VANUA LEVU       3513     3226     6121              4424

VITI LEVU        4977     2004     5858              5133

ALL              4208     2249     5907     3688     4716

_________________________________________________________

Table * Average Registered Voters Per Seat (by ethnicity and region)

_________________________________________________________

FIJIAN   GENERAL  INDIAN   ROTUMA      ALL

_________________________________________________________

ROTUMA                                      3688     3688

EAST             3133     3226     6121              3884

CENTRAL          4513     2675     5635              4725

WEST             5518     1334     5985              5442

ALL              4208     2249     5907     3688     4716

_________________________________________________________

APPENDIX *     BASIC POPULATION TABLES

  Table *   Basic Population Parameters (1956-76)

_______________________________________________

1956        1966       1976

_______________________________________________

T.Fer.R.

Fij        5.6         4.3         4.1

Ind          5.6         3.7         2.8

Inf.M.R.

Fij              50.0        36.5        23.5

Ind          79.0        44.0        29.0

Life Exp

Fij              62.3        64.2

Ind                61.0        62.4

_______________________________________________

[Source: Seniloli (1993, Table F1)

Table *  Seniloli’s Medium Variant Projection Using 1986 Base  

_________________________________________________________

Actual                  Projection          

1991        1991   1996   2001   2006   2011

_________________________________________________________

Total       746.3       744.0  800.0  857.0  915.0  976.0

Fijians     368.7       366.0  409.0  456.0  504.0  553.0

Indians     340.7       346.0  356.0  365.0  373.0  383.0

Others       36.9        32.0   35.0   36.0   38.0   40.0

_________________________________________________________

Percentages

Fijians     49.4        49.2   51.1   53.2   55.1   56.7

Indians     45.6        46.5   44.5   42.6   40.8   39.2

Others       5.0         4.3    4.4    4.2    4.2    4.1

Total        100         100    100    100    100    100

________________________________________________________

 

  Table *   Seniloli’s Estimate of Over 20 Population (Medium Variant)

____________________________________________________________________

1986          1991   1996   2001   2006   2011

____________________________________________________________________

Fijians > 20         159.1         181.3  207.3  238.7  274.4  310.8

Indians > 20         181.5         180.6  195.5  216.0  237.0  251.6

Others > 20           15.6           5.3   20.3   14.2   18.8   21.1

Total > 20           356.3         367.3  423.1  469.0  530.2  583.5

____________________________________________________________________

% Fij              44.7          49.4   49.0   50.9   51.8   53.3

%Ind              51.0          49.2   46.2   46.1   44.7   43.1

% Oth               4.4           1.5    4.8    3.0    3.5    3.6    ____________________________________________________________________

[Source: Seniloli (1993) Tables F5, F9]

Note:   Odd figures for “Others” for 1991 indicates some major problems somewhere?

  Table *   Bureau Projection of Over 20 Population (Medium Variant)

_____________________________________________________________________

1986           1991   1996   2001   2006   2011

_____________________________________________________________________

Fij > 20              159.3         181.4  206.6  236.3  270.4  305.8

Ind> 20              173.9         180.1  194.2  214.8  235.9  252.3

Oth > 20               19.3          20.0   22.0   24.5   27.0   29.5

Total > 20            352.4         381.5  422.8  475.6  533.3  587.6

_____________________________________________________________________

% Fij               45.2          47.5   48.9   49.7   50.7   52.0

% Indo              49.3          47.2   45.9   45.2   44.2   42.9

% Others             5.5           5.2    5.2    5.2    5.1    5.0

_____________________________________________________________________

Note: Indian medium projection higher than 1991 actual by 4700.

Note: Fijian Medium projection lower than 1991 actual by 2500

Note: Others Medium projection higher than 1991 actual by 900

  Table *   Bureau’s Projection of Over 20 Pop (Medium Variant) With Simple Proportional Adjustment for 1991 Error

_____________________________________________________________________

1986          1991   1996   2001   2006   2011

_____________________________________________________________________

Fij > 20                            183.9  209.4  239.6  274.2  310.0

Ind> 20                            175.4  189.2  209.2  229.7  245.8

Oth > 20                             20.0   22.0   24.5   27.0   29.5

Total > 20                          379.3  420.6  473.3  530.9  585.3

_____________________________________________________________________

% Fij               45.2          48.5   49.8   50.6   51.6   53.0

% Indo              49.3          46.2   45.0   44.2   43.3   42.0

% Others             5.5           5.3    5.2    5.2    5.1    5.0

_____________________________________________________________________

 

Table *     Expected Number of Seats in 70 Seat Parliament

______________________________________________________

1996     2001   2006   2011

______________________________________________________

Seniloli’s Estimates

Fijians                      34     36    36

Indians                      32     32    31

Others                        4      2      3

Bureau Estimates

Fijians                     34       35     35     36

Indo‑F                      32       32     31     30

Others                       4        4      4      4

Adjusted Bureau Estimates

Fijians                     35       35     36     37

Indo‑F                      31       31     30     29

Others                       4        4      4      4

_____________________________________________________

    Table *  Differences in Assumptions of Life Expectancy at Birth

____________________________________________

86‑90  91‑95  96‑00  01‑05  06‑10

____________________________________________

SENILOLI

Fij/Male   64.3   65.5   66.5   67.3   68.1

Fij/Fem    66.0   67.7   68.4   69.1   70.8

Ind/Male   60.0   60.5   61.0   61.5   62.0

Ind/Fem    65.9   67.0   68.0   69.9   70.1

BUREAU

Fij/Male   63.6   64.6   65.6   66.5   67.5

Fij/Fem    66.0   67.3   68.7   70.0   71.3

Ind/Male   60.2   61.5   62.8   64.1   65.4

Ind/Fem    65.6   66.6   67.6   68.5   69.5

DIFFERENCES (SENILOLI ‑ BUREAU)

Fij/Male    0.7    0.9    0.9    0.8    0.6

Fij/Fem     0.0    0.4   ‑0.3   ‑0.9   ‑0.5

Ind/Male   ‑0.2   ‑1.0   ‑1.8   ‑2.6   ‑3.4

Ind/Fem     0.3    0.4    0.4    1.4    0.6

___________________________________________

 

 

Table *  Differences in Assumptions on TFR

____________________________________________

86‑90  91‑95  96‑00  01‑05  06‑10

____________________________________________

Seniloli High Variant

Fij         4.1    4.1    4.1    4.1    4.1

Indo        2.7    2.6    2.5    2.4    2.3

Bureau High Variant

Fij         4.1    4.1    4.1    4.1    4.1

Indo        2.8    2.8    2.8    2.8    2.8

Difference: (Seniloli ‑ Bureau)

Fij         0.0    0.0    0.0    0.0    0.0

Indo       ‑0.1   ‑0.2   ‑0.3   ‑0.4   ‑0.5

___________________________________________

Seniloli Medium Variant

Fij         3.9    3.7    3.5    3.3    3.1

Indo        2.6    2.4    2.2    2.0    2.0

Bureau Medium Variant

Fij         3.9    3.5    3.2    2.9    2.6

Indo        2.6    2.4    2.1    2.0    2.0

Difference  (Seniloli ‑ Bureau)

Fij         0.0    0.2    0.3    0.4    0.5

Indo        0.0    0.0    0.1    0.0    0.0

___________________________________________

 

Table *  Differences in Migration Over Period

_____________________________________________

86‑90  91‑95  96‑00  01‑05  06‑10

_____________________________________________

Seniloli High Variant (per 000)

Fij/Mal     3.3    0.0    0.0    0.0    0.0

Fij/Fem     2.1    0.0    0.0    0.0    0.0

Ind/Mal    22.1    0.0    0.0    0.0    0.0

INd/Fem    13.8    0.0    0.0    0.0    0.0

Bureau High Variant (actual numbers)

Fij/Mal    2766      0      0      0      0

Fij/Fem    1847      0      0      0      0

Ind/Mal   18224      0      0      0      0

INd/Fem   17502      0      0      0      0

Seniloli Medium Variant (per 000)

Fij/Mal     3.3    0.0    0.0    0.0    0.0

Fij/Fem     2.1    0.0    0.0    0.0    0.0

Ind/Mal    22.1   18.8   15.5   12.1    8.8

Ind/Fem    13.8   11.7    9.7    7.6    5.5

___________________________________________

Estimated from Seniloli Above

Ind/Mal* 19282  17742  16570  14671  12012

INd/Fem* 12041  11080  10350   9166   7508

__________________________________________

Bureau Estimates of Indo-Fijian Migration

Ind/Mal   18224   7399   6342   5813   5285

Ind/Fem   17502   8071   6918   6341   5765

Difference (Seniloli ‑ Bureau)

Ind/Mal    1058  10343  10228   8858   6727

INd/Fem   ‑5462   3009   3432   2825   1743

——————————————-

Note: Seniloli has higher emigrants

    Table *  Net Departure Figures (CES)

—————————————————————-

1986   1987   1988   1989   1990   1991   1992   1993

—————————————————————-

TOTAL      6590  18359  10674   9387   5805   5961   4387   5016

Fij               1105   1586    790    681   ‑101    110

Indo             14328   7254   7412   3929   4449   2990

Others            2926   1834   1185   1195   1613   1287

—————————————————————-

Actual        Seniloli

Total 1986‑1990         50815         31323

Total 1991‑96           25607         28823

Est

Total 86‑96             76422         60145

—————————————————————

Note:           Seniloli’s estimate of Indo‑F pop likely to over‑est the Indo‑F pop by some 20,000 by 1991 (does not this seem high?) and by 16,000 by 1996: for her medium variants due to emigration factor alone.

 

APPENDIX   AUTHOR’S POPULATION PROJECTIONS

Table *  Fertility Assumptions Made for My Projections

_____________________________________

89-94  94-99  99-04  04-09

_____________________________________

Fijians

TFR          3.8    3.6    3.4    3.2

GRR          1.8    1.7    1.7    1.6

MAC         28.6   28.4   28.3   28.2

Indians

TFR          2.5    2.4    2.2    2.1

GRR          1.2    1.2    1.1    1.0

MAC         27.7   27.7   27.6   27.6

_____________________________________

TFR = Total fertility rate

GRR = Gross reproduction rate

MAC = Mean age of mothers at childbirth

 

Table *  Author’s Projections For Total Population

_________________________________________________________________

1989   1991   1994   1999   2001   2004   2006   2009

_________________________________________________________________

Fijians    352.5  368.1  393.4  437.0  454.6  481.9  499.7  527.2

Indo-Fij.  338.3  340.8  344.8  348.5  349.5  349.6  348.8  347.8

Others      35.8   36.9   39.2   43.2   45.0   46.8   48.7   51.7

TOTAL      726.6  745.8  777.4  828.7  849.1  878.3  897.2  926.7

_________________________________________________________________

Percentages

Fijian        49     49     51     53     54     55     56     57

Indo-Fijian   47     46     44     42     41     40     39     38

Others         5      5      5      5      5      5      5      6

_________________________________________________________________

Seniloli            744                  857           915

_________________________________________________________________

Note: It is assumed that “Others” grow at 2% annually from 1991, although actual growth between 1989 and 1992 was only 1.5%.  This may have been counterbalanced by recent arrivals of Chinese migrants.

    Table %  Expected Population Growth Rates

—————————————–

PERIOD         Fijian       Indo-Fijian

—————————————–

1989‑1994         2.2           0.4

1994‑1999         2.1           0.2

1999‑2004         2.0           0.1

2004‑2009         1.8          ‑0.1

—————————————–

Table *  Projected Dependency Ratios (dependent persons per 100)

______________________________________________________

YEAR        <15            65 & Over     <15+ 65& Over

Fij.  Ind.       Fij.  Ind.    Fij.  Ind.

______________________________________________________

1989     67    60           7     4       73     64

1994     63    52           7     5       70     56

1999     59    42           7     5       66     47

2004     56    37           8     7       64     44

2009     52    35           8     8       61     43

______________________________________________________

Table *  Author’s Projections of the Over 20 Age Groups

_________________________________________________________________

1989   1991   1994   1999   2001   2004   2006   2009

_________________________________________________________________

FIJ > 20   174.5  184.1  200.3  228.6  240.6  261.3  274.2  294.7

IND> 20   173.5  179.2  187.6  202.5  209.4  219.5  225.1  228.8

OTH > 20    19.2   20.0   21.2   23.4   24.4   25.4   26.4   28.0

TOT >21    367.2  383.3  409.1  454.5  474.4  506.2  525.7  551.5

_________________________________________________________________

Table *     Expected Proportional Number of Seats in Parliament

_________________________________________________________________

1989   1991   1994   1999   2001   2004   2006   2009

_________________________________________________________________

FIJIANS       33     34     34     35     36     36     37     37

INDO‑FIJIA    33     33     32     31     31     30     30     29

OTHERS         4      4      4      4      4      4      4      4

TOTAL         70     70     70     70     70     70     70     70

_________________________________________________________________

Voters/seat  5.2    5.5    5.8    6.5    6.8    7.2    7.5    7.9

_________________________________________________________________

    Table *  Author’s Projections of School Age Population (000)

—————————————————————–

1989   1991   1994   1999   2001   2004   2006   2009

—————————————————————–

FIJ 6‑11    53.2   55.5   57.1   60.3   62.1   64.3   65.8   67.8

FIJ 12‑15   30.9   32.6   35.4   37.8   38.5   40.0   41.2   42.8

FIJ 16‑17   14.4   15.0   16.2   18.4   18.7   19.1   19.7   20.5

FIJ 18‑20   20.7   21.2   22.8   26.0   27.2   27.9   28.5   29.6

INDO 6‑11   52.0   51.9   47.2   39.2   37.2   35.6   34.8   34.3

INDO 12‑15  30.5   31.5   33.7   29.9   27.3   24.6   23.2   22.4

INDO 16‑17  13.8   14.2   15.0   16.3   14.9   12.9   11.9   11.1

INDO 18‑20  20.2   19.7   20.5   22.9   23.4   20.7   18.7   17.2

TOT 6‑11   105.2  107.4  104.3   99.5   99.3   99.9  100.6  102.1

TOT 12‑15   61.4   64.1   69.1   67.7   65.8   64.6   64.4   65.2

TOT 16‑17   28.2   29.2   31.2   34.7   33.6   32.0   31.6   31.6

TOT 18‑20   40.9   40.9   43.3   48.9   50.6   48.6   47.2   46.8

—————————————————————–

Note:   These do not include the “Others”.

Notes on Population Tables

*   Seniloli’s projections used the Population Council’s package FIVFIV Version 10.0, with separate projections for Indo-Fijians and Fijians, while Others assumed to follow Fijians.

*   Seniloli’s projections were based on 1986 data, not on the more recent Bureau estimates (eg the 1989 tables from Bureau or the 1992 estimates).

*   Seniloli (1993) did not use net departure figures, but those who stated they were emigrating or were going to be away for longer than a year.  This may not be the most appropriate:Bedford(1991) concluded that many who indicated short visits stayed for longer periods.

*   Seniloli assumed that the rate of migration would be 22.1 per thousand, declining gradually to 8.8 per thousand by 2010, on the grounds that “emigration could decline in response to economic circumstances and tighter emigration policies of the destination countries”.  This roughly gives annual emigration figures of 6,265 declining to 2738 by 2010.  The actual numbers emigrating suggest that these assumptions underestimate the likely rate of emigration.  Thus while Seniloli’s estimate of the 1991 total population is lower than the actual 1991 by 2.3 thousand, the Indo-Fijian population is over-estimated by 5.3 thousands, while the Fijian population is under-estimated by 2.7 thousands and Others underestimated by 4.9 thousands.  In fact her estimates of “Others” seems seriously out everywhere.  These differences could significantly expand by the end of the projection period.

*   Why are female emigrants higher than male emigrants in the Bureau’s estimates?

*   According to the Bureau, amongst the net resident departures every year (as recorded by the CES) are some 1000 or so persons who came in as visitors and later changed their status to “residents” after they acquired their work permits.  i.e. the total net departures need to be reduced by the equivalent number.  But these are to be found predominantly amongst the “other” categories.

*   The 1989 Population Projections made in the Bureau of Statistics are significantly different from the Bureau’s estimates for 1991, although the High Variant projections for Fijians are almost exactly correct.  Those for Indo-Fijians are significantly out, probably because of the migration assumptions.

Advertisements

Comments are closed.

%d bloggers like this: