Interview: Manoj Gokhool – Chairman of International Centre for Electoral Psychology
… and take the risk of generating significant democratic crises”
“The selective concept of wasted votes means that those who win in small constituencies and lose in big ones get more votes counted under the List system than those who win in large constituencies and lose in small ones!”
“History has shown that if politicians want to be trusted, they have to trust the electorate”
Manoj Gokhool, Chairman of the UK-based International Centre for Electoral Psychology, is considered as one of the foremost experts in voters’ psychology and the determination of choice. He has given several lectures and speeches worldwide on the subject of choice and sensation transference. He recently ran a large-scale electoral psychology survey in Georgia.
He has worked with a number of Heads of Governments and Electoral Commissions on Civic Engagement and Democratisation. He is presently the Chief Executive of the International Centre for Parliamentary Studies. He was a founder-member of Parliamentary Communications, the largest parliamentary and political communication organisation in Europe. He is also on the BBC panel of experts on Electoral and Political affairs and has made various TV appearances on political programmes. He was the opening speaker introducing the House of Commons Speaker’s Inaugural Lecture on Reforming Parliament: New Parliament New Opportunity. Manoj Gokhool is presently spearheading a comprehensive independent review of the Structure of Governance in the UK covering Westminster, Whitehall, Local Government, Media, Trade Union influence, Party Funding and Voters Relationship.
We have sought his views on the Rama Sithanen-inspired electoral reform proposals of the Labour Party-MMM alliance and their likely impact on future electoral outcomes. His comments confirm this paper’s earlier stand that we have to proceed with caution when it comes to overhauling the electoral system…
Mauritius Times: In your experience, what have been the unintended consequences of electoral reform in other countries and what do we have to guard against? Are, for instance, the simulations based on the past electoral results prior to introduction of the reforms reliable?
Manoj Gokhool: My experience is that the experience of other countries is simply not good enough, and it is crucial to simulate potential effects. Simulations are reliable only if they are done in the most professional and rigorous manner. This is not something that you can improvise and there is a big science behind understanding the consequences of electoral systems. They also should not just look at past election results but at theoretical problem cases too as the past is not always a comprehensive indication of what might happen in the future. So conducting independent evaluations is essential.
* Here in Mauritius the Labour Party and the MMM have agreed on an electoral programme based mainly on two underlying principles: maintenance of the First Past the Post system coupled with the introduction of 20 Party List seats wherein the wasted votes system would come into play, and sharing of power by the President of the Republic and the Prime Minister in the context of a Second Republic. Does such a hybrid system exist anywhere else in the world and has it promoted democracy, stability, etc?
A lot of hybrid systems exist – for instance in Germany, Hungary, etc. However, I am not aware of another system which conceptualises ‘wasted votes’ in such a narrow way… Typically, political parties will have their own proposals evaluated by independent specialists and will also give several options to voters. This was the case in two recent referenda on new electoral systems in the UK and New Zealand for instance (2011).
As for power sharing between the President and the Prime Minister, it is not only about the safeguards in the Constitution that makes such a system effective, it is also the politics behind it all. In fact constitutional safeguards can be used as political tools either to paralyse the proper working of the government machinery or even as a usurpation tool against democracy. Redistricting Laws in the US are good examples of the latter.
At the end of the day, an effective government depends on its workability which in turn depends on the personalities and politics behind it all.
I do a lot of work in the former Soviet Union where I am running several electoral psychology projects and where democracy is in an exceptionally dynamic state. Look at what has been happening in Georgia this week: the Constitution was recently successfully amended to offer power sharing. However, the problem did not arise from constitutional deficiencies but from tensions between personalities and political interests.
The Constitution can provide a framework but cannot make a government work or people work harmoniously together. Most of the proposals that came from the President of Georgia were being systematically rejected by Parliament, which affected the functioning of the government in Georgia.
However, it will be also wrong to assume that the leaders of the Labour Party and that of the MMM will not be able to work together and that it is all bound to fail. Change always brings discomfort and if the two leaders really believe in what they are doing, they will articulate it and get the buy-in of the population.
* The Electoral Reform Consultation Paper outlining the Mauritian government’s stance on electoral reform provides for additional seats to be allocated based on a wasted votes system? What are your views on the method being proposed?
These refer to the proposals of Rama Sithanen. In my opinion, his proposals have been the best so far. However, I have two main reservations: firstly, the ‘Unreturned Votes Elect’ (UVE) formula is only taking into consideration one type of wasted votes i.e. losers’ votes and, secondly, and more importantly, the winners in the larger constituencies are penalised twice (this is because of acute malapportionment in Mauritius). Let me illustrate this by a very simple example:
To keep things simple, let us assume that we only have two constituencies and that one has twice the population of the other, that one party is stronger in the larger constituency while the other party is stronger in the smaller one. So we have:
Constituency 1: 100,000 voters, 3 seats
Party A: 60% (60,000 votes) = 3 seats
Party B: 40% (40,000 votes) = 0 seats
Constituency 2: 50,000 voters, 3 seats
Party A: 40% (20,000 votes) = 0 seats
Party B: 60% (30,000 votes) = 3 seats
Altogether, Party A won 80,000 (53.3%) votes and Party B 70,000 votes (46.7%). Party A is thus the clear winner of the popular vote and should really have the majority of the seats.
With the current (uncorrected) system, this is not what is happening. Because of the differences in constituency sizes, in effect, the two parties get 3 seats each so it is a tie between them in Parliament.
With the proposed reform, however, the votes of the losing party in each constituency are considered wasted and attributed proportionally according to a list system.
This means that with the new system, it would be considered that Party B has 40,000 wasted votes in Constituency 1, while Party A has 20,000 wasted votes in Constituency 2. In other words, if there are 3 seats to attribute through the List system, Party B – the electorally weaker of the two parties — will in fact get two of them while Party A, which has won the popular suffrage, will only get 1!!
So with the new system, instead of a tie, the proposed correction creates an even more unfair result: Party A wins 53.3% of the votes but only 4 out of 9 seats (44.4%), and Party B has clearly lost the election in terms of votes (46.7%) but is now made to clearly win the elections in terms of seats obtained with 5 out of 9 seats!! (55.5%), as illustrated in the Table below:
In effect, Party A is punished twice for doing well in larger constituencies: firstly, because its voters need more votes than Party B voters to be represented in the first place; secondly, because the selective concept of wasted votes means that those who win in small constituencies and lose in big ones get more votes counted under the List system than those who win in large constituencies and lose in small ones! This is a ’double peine’ for the parties which win the popular vote!
On an associated issue, the electoral reform proposals suggest that having only 20 seats allocated by Proportional Representation would limit the risk of weakened majorities. However, it is essential to note that, according to this system, elected candidates prevent their party from competing for the 20 PR seats as illustrated below:
– If a party overwhelmingly dominates the election, the system will force the 20 seats in the hands of the opposition. Thus, for example, if across all constituencies, Party A wins 89% of the votes and all constituency seats, Party B wins 10%, and Parties C and D share the last 1%, the UVE system will force 20% of the seats in the hands of Party B as Party A will be disallowed from competing for those seats.
Even more problematically, if the election returns a very narrow result between Parties A (ahead) and B (behind) but one party is equally strong across all constituencies while the other tends to be strong in some but weak in others, then the likely result is that Party B could end up with significantly more seats in Parliament than Party A despite receiving fewer votes.
This is why wasted votes should be defined (as is more common in political science) to also include all the votes that the winner did not really need to win (this is, for example, illustrated by the literature on malapportionment and gerrymandering such as Erikson, R. 1972. – ‘Malapportionment, Gerrymandering, and Party Fortunes In Congressional Elections’ in American Political Science Review.
I have to stress at this point that it is essential that wasted votes be dealt with as indeed they mean that some people are represented while others are not, and when numbers pile up, this can make for a very biased representative system.
The problem with the Electoral Reform Consultation Paper is that while political science knows of several essential types of wasted votes, the Consultation Paper only considers one of them. This is a critical problem because while ignoring wasted votes is unfair, acknowledging some wasted votes while ignoring others which may potentially be even more important in numbers could result in a system which might not be less but in fact even more unfair, especially as our constituencies tend to vary a lot in size! This needs to be carefully modelled so that the Mauritian people know exactly how the proposed consideration of wasted votes might improve or worsen the situation and how it could be made fairer.
As I have mentioned from the outset, there is no perfect electoral system and any system requires some sort of trade-offs. The system proposed by Rama Sithanen also implies some specific trade-offs, which he and maybe many others consider as acceptable. It would be important to find out whether the majority of the population also considers these trade-offs acceptable.
* In other words, what you are saying is that given that Mauritius has unequal constituencies and the smaller constituencies tend to support a particular party and the larger ones support another party, could one party have more wasted votes than the other and hence more of the additional seats inspite of having lost the elections nationwide?
Yes, but again, this is mostly a problem in that in the proposed system not all wasted votes are taken into account at all! So yes, as a result, this creates a very big risk of that kind.
* If the proposed electoral reform providing for additional PR seats (based on the wasted votes system) were to be enacted, can we say at this stage how this particular reform would impact the electoral outcome?
No, again, you don’t just put an approximate answer together and hope this may be approximately true! You need independent specialists to tell you how this might work including in both best-case and worst-case scenarios, and evaluate the opportunity to consider small modifications to the proposed system so that it works better and/or more reliably. This is an indispensable step that cannot be avoided, because if you cut corners and say, ‘I think that this will be the effect’, then you are playing Russian roulette with your representative system and taking the risk of generating significant democratic crises. The effects may be very good – or not, this must be assessed rigorously by professionals
* Let’s be more specific: we may perhaps reliably guess as to what the overall outcome would be when two major parties come together to contest the elections against an alliance of smaller parties. What if these two major parties (Labour Party & MMM) are in opposition to each other and one of the two (say, the MMM) enjoying the support of a smaller party (the MSM, for instance)? What do simulations point to?
No, we can’t guess. We need the simulations done specifically.
* There are what one could qualify as particular political circumstances presently prevailing in Mauritius. The buzz in different sections of Mauritius society but mostly amongst the traditional vote bank of the Labour Party would suggest that the latter electorate would not be comfortable with their Party’s alliance with the MMM. We cannot say at this stage whether electoral arithmetic would still prevail, and how much of the LP’s core electorate would nevertheless support their Party. What do (your) simulations suggest in the light of the 50/50 sharing of tickets between the LP and the MMM should the MSM-led alliance succeed in making a dent in LP’s electoral bank?
There can be various reasons for such discomfort on the part of both vote banks but one should never forget that both the MMM and the Labour Party have deep roots and their respective vote banks have a sense of identity that is directly associated with their respective parties. We know from our various works on voting that, at the time of voting, the decisive and most important element is identity. In principle, a MMM-Labour Party coalition is bound to win the election. It would, however, be interesting to find what constitutes the elements in each of the parties that their respective electorates identify themselves with and also whether these elements are also present in the other political parties i.e. are there emotional alternatives?
Last week I was doing some work in Washington for the mid-term elections. I met an Afro-American lady outside a polling station and asked her how was the election going. She had lots of complaints about the Democrats and Obama’s administration. So I said to her, “I take it you voted Republican.” Her answer was spontaneous. “Of course not,” she said, “my parents would turn in their graves if I were to vote Republican.”
It was clear from her answer that however uncomfortable she was with the Democrats, at the point of voting, she could only identify herself with the Democrats be it for historical or perceived historical reasons.
Recent studies have shown that 29% of Americans and 40% of French have changed their voting decisions on the day of the elections – and this is not predominantly because they have been convinced to change their minds but rather because while they were swinging weeks before the elections, at the decisive moment, they voted for whatever constituted their identity. The Mauritian voters are not very different, and this is why I say, parties with deep roots, entrenched in unique identity platforms tend to reap consistent results until these identity platforms have been dissolved or shaken through generational or societal developments.
* In general terms, what are the circumstances and reasons that would otherwise lead a country to change its electoral system and adopt a new one?
Typically, electoral systems are changed because something is not working with the current one. For instance, it may produce too much instability, or it may be seen as insufficiently representative, meaning that some citizens or regions are over-represented compared to others, or it could foster division rather than unity in the country, or it may not be fit with the modernisation of society. It is also the case that democracies ‘age’, and one would have different priorities and safeguards in the context of a democratising country as compared to when democracy is more mature.
It is also important to note that no electoral system is perfect. Any electoral system implies some trade-offs and choices that need to be made carefully and consciously, evaluated independently so that one can think not only of what a new electoral system could improve, but also possible shortcomings and what would be the ‘worst case scenario’ that citizens should have in mind when deciding whether to support a change of system or not.
One should bear in mind that that elections are not only there to transform votes into seats but also to cement citizens and give them a sense of ownership, responsibility, and efficacy within their political systems.
* Once there is a national consensus on the need to change the electoral system, how does a country decide on which new electoral system would best suit the country’s requirements?
There is no hard and fast rule or one-size-fits-all system. However, typically, politicians would make proposals and try to reach a consensus, then they would organise an independent evaluation of the proposals to assess the likely consequences of the proposed changes; they would next discuss independent evaluations and either update the proposals accordingly or submit it to citizens’ vote together with the evaluations that have been conducted. This is the most democratic way to proceed. This is a crucial part of democratic accountability and transparency. It is crucial that nobody be judge and party.
* What are the criteria that are normally used to assess the fairness of an electoral system?
Fairness is just one element be it a very important one. However, there are some other equally important elements, and these relate to the institutional framework, i.e. Government Stability (including, in our view, the emergence of clear majority), Effectiveness (in enabling governmental alternance if so wished by voters), Accountability, Avoidance of Communitarianism, as well as Socio-demographic Inclusiveness and Gender Representation.
There are also non-institutional elements which can be classified as Civic and Psychological. They include Readability (sufficiently straightforward and readable so that citizens can relate to it); Psychological Inclusiveness (all citizens should be in a position to feel that they can relate to the logic of their electoral system regardless of their background, community, or education); Ownership (electoral reform should be seen as serving the interest of citizens); Engagement (a good electoral system must make citizens want to participate in elections and feel that their participation matters both symbolically and politically); Identity Consolidation (elections should consolidate citizens’ political identity and their perspective of ‘sharing’ important moments in the life of their country); Continuity (an extremely radical electoral system reform sends a very different signal to citizens and parties alike than a reform which, however significant, appears to be rooted in an existing tradition. Indeed, creating ‘significant losers’ — be it amongst parties or citizens — can lead to significant backlash and a risk to deepen some social, political, or ethnic fragilities which the system should, instead, try to mend).
One might say that it all looks too complicated and it should be left to specialists to deal with these matters and the average voter would not understand it anyway. In my opinion, as we are dealing with the most precious asset that any citizen possesses in a democratic society, however complex these issues may be, it is up to the specialists and political class to ensure that the choices are clearly understood so that voters can make informed choices. Ultimately, history has shown that if politicians want to be trusted, they have to trust the electorate.
I would also like to point out that Mauritius has a highly respected independent Electoral Commission which can play an important role in that process.
* How does the buy-in of the people through a referendum affect the way the new electoral system is ultimately perceived and applied?
Asking the population to decide makes them more likely to accept the outcome of the new electoral system should they decide to vote for a change. In effect it also makes the voters accountable.
Let us say, for example, that there has been no consultation with the voters. It is highly likely that after the first election under the new electoral system, the losing party/parties would do a simulation of what the result would have been had there been no change. If the outcome would have been different, this could be used to cast doubts on the legitimacy of the newly elected government. On the other hand, if they have been consulted, they might not like the result, but they are more likely to accept it because they have been part of the process.
* What do you consider as the most important future development in the democratic cycle of Mauritius?
The development of the next generation of political leaders.
* Published in print edition on 14 November 2017