Skip to content

“The unfairness of the Modified D’Hondt electoral system in Fiji” (FT 24/11/2018)


“The Unfairness of the Modified D’Hondt electoral system in Fiji”
(edited version in The Fiji Times 24/11/2018)
Professor Wadan Narsey

For years now, claims have been made that the electoral system in the 2013 Constitution and the Electoral Act 2014 is the best ever and the most proportional, where all votes have equal value.

While I have always argued that the massive 5% threshold has totally destroyed the value of a large numbers of votes which were cast for independents and small parties (some 27,000 votes in the 2014 Election and about 11,000 in the 2018 Election) even I was under the impression that the modified D’Hondt system used by the Fijian Elections Office under the Electoral Act was was still largely proportional.

But I have been proven wrong by the 2018 Election results declared by the Fijian Elections Office.

FFP (with 227,241 votes) has been awarded 27 seats, SODELPA (with 181,072 votes) has been awarded 21 seats and NFP with 33,515 votes, awarded 3 seats- all correctly according to the 2013 Constitution.

But I suggest here that it would have been much “fairer” by any sense of the word to award NFP one more seat and FFP one less, which would reduce the majority of FFP in Parliament to only 1- quite a decisive correction.

Indeed, NZ’s proportional system, internationally accepted, would have allocated FFP 26 seats (one seat less) and NFP 4 (one seat more than currently).

Even a primary school child can understand my arithmetic logic below.

All apparently lawful: according to the Modified D’Hondt system

Assuming that there are no substantial irregularities as claimed by the Opposition Parties, then the above results as declared by the FEO are indeed according to the electoral vote counting  clauses in the Electoral Act  under which all the parties contested the 2014 and 2018 Elections- the modified D’Hondt system.

The modified D’Hondt system is extremely difficult to explain to ordinary people, but let me give it a go, just using FFP and NFP in the example, to simply the calculations.

(a) there are 51 seats to allocate to parliament.

(b) divide each qualifying party’s total votes received by 1, 2, 3, etc. going down to 51.

(c) the Fijian Elections Office chooses the MPs from all the parties, with the highest average number of votes.

(d)        So if only 1 FFP MP was elected, that MP would represent 227,241 voters.

if 2 were elected each MP would represent 113,621 each (or 227,241/2)

Keep dividing by increasing numbers of MPs until you get to 26.

if 26 were elected these FFP MPs would each represent 8740 votes each (or 227,241/26).

(e)        Now if only 1 NFP MP was elected, that MP would represent 33,515

if only 3 NFP MPs were elected they would each represent on average 11,172 voters

(f) by now 21 SODELPA MPs have also been selected: so the total number of MPs already selected for parliament = 26 +21 + 3  = 50.  One more to go.

(g) The FEO now has to select the 51st MP and the following are the average number of voters that each party would have, if one extra MP from that party was selected.

Average voters

27th FFP candidate:                  8416

22nd  SODELPA                        8231

4th NFP candidate                    8379

(h) You can see that the modified D’Hondt rule as in the Electoral Act requires the FEO to choose the next MP who would give his or her party the higher average number of voters our of these three parties.

Clearly, in this case, the 51st MP will come from the FFP, because the 27th FFP MP will give an overall average of 8416 voters, still higher than the other two parties’ candidates should they be selected.

So if the numbers of votes are correct as stated, then the FEO has followed the law as it stands in the Electoral Act 2014

But is it fair in the common sense of the word?

Especially when the Fiji voting public have been led to believe for five years that every vote should have equal value?

What if there was strict proportionality?

If one accepts the 5% threshold which eliminates the small parties, then the qualifying parties are as follows, with their strictly proportional share of votes and their ideal allocation of seats.

By strict proportionality,

FFP is entitled to 26.2 MPs,

SODELPA            20.9 MPs

and NFP               3.9 MPs.






There is little to quibble about SODELPA getting 21 seats (which is 20.9 rounded up to 21).

But then why is the NFP entitlement of 3.9 not rounded up to 4.

Why is FFP entitled to only 26.2  getting 27 seats when common sense (and any primary student) would know that the 26.2 should be “rounded down” to 26 as universally accepted.

This question is even more valid when one notes that the current FEO allocation, while according to the law, results in some MPs representing far more voters on average than others, a result severely criticized in previous electoral systems Fiji has had.

How many voters does an MP represent?

Remember the many debates we used to have twenty years ago about previous electoral systems where it was alleged that it was unfair that some ethnic groups or some rural constituencies effectively and unfairly represented much fewer or much larger numbers of voters than others? It was then alleged that the votes were therefore not of “equal value”?

Of course, such equality of voters represented by MPs cannot be exactly equal in practice, and some deviation is to be expected.  But how much deviation is reasonable? And how much unreasonable?

Look at the current situation as given by Table 2 (forgetting for the moment all the 11,000 voters nullified by the 5% threshold rule:.





Each of the 51 MPs represents 8,663 on average.

The 27 FFP MPs represent 8,416 votes on average, or a small 3% below the national average.

The 21 SODELPA MPs are pretty close to the national average.

BUT the 3 NFP MPs represent a massive 11,172 voters on average OR  29% MORE than the national average.

Common sense would suggest that this is not fair.

Indeed, Table 3 gives you the averages and more acceptable deviations, if FFP was reduced to 26 seats and NFP was increased to 4.





Now the differences from the average are far more reasonable for ALL parties, including FFP and NFP.

This is of course what would be indicated by strict proportionality.

Fiji does not need the Modified D’Hondt system at all.

Why is the D’Hondt system unfair?

Some experts say that a D’Hondt system (whether the standard or modified version)  is unfair to small parties. Possibly that is because the averages fall very quickly when the party is small (like NFP with less than 5 probably entitled) but the decreases in averages are much smaller for larger parties, like FFP and SODELPA, when you get to 20 or more seats.

But note that even in the current 2018 situation, the result could have been fair to NFP for example, if they had received only 132 votes more (with 132 less for FFP).

My calculations indicate that in that case, the FEO would have would have selected the 4th NFP candidate to be  the 51st MP to the Fiji parliament. But that does not take away the unfairness of the current situation.

Let us remember that votes to FFP declined from 2014 by 23%, while that to the small NFP party, increased by 24%.

How can then a 0.2 seat for FFP become 1 while a 0.9 seat for NFP becomes zero.

Hardly the kind of “rounding up” arithmetic one wants to teach our school children.

Indeed, the current practice may not be following the spirit of what is in the 2013 Constitution itself.

Does Fiji’s Counting Practice follow the Spirit of the 2013 Constitution?

We now have a 2013 Constitution and electoral system which has been imposed on the people of Fiji, with some strengths (like broad proportionality) but still many negative aspects, some of which I have pointed out in this article such as the unfair 5% threshold rule and the modified D’Hondt system which can be, and has been unfair in the 2018 elections.

But note that that all the 2013 Constitution (Section 53) requires is that

(a) a party’s allocation of seats in parliament must be in proportion to the votes received and that

(b) “a written law shall make provisions” including “prescribing rules for awarding seats … that accord with an internationally accepted method for awarding seats to candidates within an open list system of proportional representation”. (my emphasis).

While the 2013 Constitution does not specify any “internationally accepted method” let us go to our nearest neighbor NZ, who contributed a large number of observers to the recent MOG (and also gives Fiji lots of money and jobs).

The much fairer NZ system

The NZ reps to the MOG jaunt in Fiji, could have pointed out for instance, that according to NZ’s proportional system, the votes received by FFP, SODELPA and NFP would have entitled them to 26, 21 and 4 seats respectively.

Readers might want to look at the NZ Elections website which even has a “MMP Seat Allocation Calculator” which will give you the above results if you input the Fiji elections results for FFP, SODELPA and NFP.

Click “count of Party votes”  at the top row.

Input the numbers of votes in the first three rows for FFP (227241) , SODELPA (181072) and NFP (33515)

Click the orange button at the bottom (“Calculate Parliamentary Seats”)

You will get the number of seats in the NZ parliament (120 seats altogether) for these parties:   62, 49 and 9 respectively

Convert proportionally to Fiji’s 51 seat parliament you get 26, 21 and 4.

This was indeed the system that I had recommend to the Yash Ghai Commission and a workshop at USP which put my proposal into practice: a proportional system with local constituencies.

Incidentally, if local constituencies were in existence, and you put in the number won by each party, the NZ Elections Office Calculator would also estimate how many “List” MPs would be selected into to ensure a truly proportional result overall.

And if there was a “Closed List” with alternating women candidates, then you would also guarantee a larger proportion of women MPs.

What to do?

Given the closeness of the 2018 Election Results, the apparent willingness of the Opposition Parties to co-operate with the Bainimarama Government, and the many calls from the public at large for such co-operation, I would suggest that Parliament (i.e. FFP) needs to set up a Constitution Review Committee which would urgently examine all the areas of the 2013 Constitution and the Electoral Act over which the Opposition parties are voicing concern.

Of course, this is unlikely to happen and progressive forces in Fiji will need to once again struggle to make it happen. What a pain in the butt.

All this reminds me of what many of us faced after the 1990 Constitution was imposed on Fiji by the Rabuka Government after the 1987 coup.

Many progressive academics and NGOs entered into dialogue for several years and contributed to the deliberations of the Reeves Commission, whose report then formed the basis of the 1997 Constitution (plus the sadly forgotten multi-party cabinet provision) approved unanimously by the Fiji Parliament.

Then there was a Yash Ghai Constitution Commission, which we all similarly contributed to but its report was sadly rejected by the Bainimarama Government.

Today, not only is there the unfairness I have pointed out above, but there are other undesirable results of the unilaterally imposed 2013 Constitution and Electoral system:

  • For instance all Fiji must wonder that in the 2014 and 2018 Elections,  any number of candidates with large numbers of votes fail to get elected, while many are “elected” into parliament with a pitiful small number of votes, riding on the coat-tails of some Leader.
  • Or not having local constituencies with local candidates who voters can identify with and who can be held directly accountable for local issues.
  • Or having a very unfriendly ballot paper with names, numbers and photos but not party symbols.
  • Or the many other aspects of the imposed 2013 Constitution which sensible people committed to human rights and freedom of the media find quite objectionable.

Of course, many of the older cynics (myself included) might understandably protest at yet another grand exercise likely to end in a waste tax-payers’ funds and the valuable time of public intellectuals and NGOs, as had happened so casually with the Yash Ghai Commission exercise.

But the younger generation who have not gone through such exercises before, might still have the energy, even if we old dogs do not, to tackle the huge problems because “the times they are a’changing” and have changed for the worse.

They better do so, for they are the ones who will be paying for Fiji’s massive public debt.


Comments are closed.

%d bloggers like this: