2023-4 The Netherlands

To be published in MSSR (Munich Social Science Review)

Reviewing Dutch Democracy 

Peter Emerson 

Abstract: In the current review of the Dutch political system, there is one huge  'elephant in the room' which has so far not been questioned: the fact that in the  Netherlands (and elsewhere), the chosen methodology of decision-making in  Parliament is the simple binary vote. So while arguments about numerous  electoral systems abound, discussions of the various methodologies which could be  used to identify the democratic ‘majority’ opinion do not. This is despite the fact  that the current methodology, majority voting, is primitive, often divisive, and  sometimes hopelessly inaccurate. More inclusive methodologies are possible, the  use of which could lead to a more wholesome polity.  

Furthermore, one of the more accurate – and therefore more democratic – decision-making methodologies is non-majoritarian; if the latter were to be  adopted, the ubiquitous habit of dividing every newly elected parliament into two  could be replaced by a more inclusive form of governance based on an all-party  coalition. Hence, and at a stroke, many of the current problems associated with  forming a new government could be overcome… and the benefits abroad could be  overwhelming! 

Keywords: Preferential voting in decision-making, all-party power-sharing,  modified Borda count, governments of national unity, matrix vote, conflict  resolution. 

1. Introduction 

In reviewing the Dutch system of governance, the first four papers in the  Munich Social Science Review (MSSR), Volume 3 (2020) discussed the  Dutch electoral system, even though, as Hannu Nurmi (2020) implies, 'if it  ain’t wrong, don’t fix it.’ In contrast, this paper concentrates on two 

© 2023 Verlag Holler, München. 

ISSN 0170-2521  

ISBN 978-3-88278-315-5 

www.accedoverlag.de 

38 Munich Social Science Review, New Series, vol. 6, 2023 

aspects which do need fixing: the question of how best should decisions be taken, that is, how best can be identified the will of parliament; and the  second, how best to implement government formation. A solution to the  first could well facilitate a resolution of the second. 

Now democracy is based on the principle of majority rule. That is not  in doubt. The question is: which voting procedure should be used to  identify the democratic majority opinion? Apart from binary voting, there  are a number of multi-option procedures: these include  

(a) plurality voting;  

(b) the two-round system TRS;  

(c) the alternative vote AV;1 

(d) approval voting; (e) range voting;  

(f) the Borda count BC, 

(g) its original, the modified Borda count MBC; and  

(h) the Condorcet rule.2  

While electoral systems vary enormously, decision-making does not, and  the elected parliaments of most countries nearly always use binary voting.  There are just a few exceptions. Denmark uses plurality voting, albeit on  only three options. Norway tried TRS, but only once. Sweden has serial  voting – a series of majority votes which can give the Condorcet winner – when discussing amendments. And still in Scandinavia, Finland has used  TRS in a referendum. Interestingly enough, so too has the UK.3 Indeed, it  is in referendums that some countries have allowed for a degree of  pluralism seldom if ever seen in their parliaments;4in 1992, for example,  New Zealand held a five-option TRS referendum on its electoral system.  

Of the voting procedures listed above, the MBC is non-majoritarian: it  can identify the option with the highest average preference, and an  average, of course, involves every member of parliament, not just a  majority of them. If this were to be the basis of Dutch democracy,  

1 Otherwise known as the single transferable vote STV; or, in the USA, as ranked  choice voting RCV; or again, in Australasia, as preference voting PV. 2 Apart from binary (yes/no) voting, the only other decision-making voting  procedure mentioned in the Munich Social Science Review (MSSR), Volume 3, was approval voting (in Brouwer and Staal 2020). 

3In 1948, the British Government proposed a binary referendum for New foundland. The folks in Halifax were soon on the streets to demand a third option;  it was duly added… and on an 88% turnout, it won the second round by 52%. 4 The Netherlands has not used the referendum tool very often. But uniquely, in  1991, the good people of Vlaardingen held a referendum on whether or not to have  a referendum.

P. Emerson: Reviewing Dutch Democracy 39 

governance could be based on an all-party,5 power-sharing government of  national unity GNU. 

Accordingly, this paper will consider the weaknesses of binary voting  before then describing the MBC, its procedures and its advantages. It next  discusses an outline of the matrix vote, a tabular voting procedure by which members of a parliament can choose, in their order of preference,  not only those whom they wish to be in cabinet, but also the particular  ministerial post in which they wish each of their nominees to serve. The  paper concludes with the potential benefits of consensus voting. 

2 Binary Voting 

Countless incidents show that binary voting is inadequate. Sometimes,  indeed, the outcome of such a decision-making procedure is determined by  a totally unrelated incident. In 1900, in a debate on compulsory education,  F.D. Graaf Schimmelpenninck fell off his horse and missed the vote,  which thus was approved by 50:49. A more recent incident was in 1999 – De nacht van Wiegel – when the Government collapsed as a result of the  said Mr. Wiegel, a member of the Eerste Kamer or Senate, voting against  his own Party, Volkspartij voor Vrijheid en Democratie VVD. In both  instances, the matter was resolved by a margin of just one vote. Nearly  every democracy has witnessed similar events in which the course of  history has been altered, completely, by the act of just one person – in  some cases, an individual who was bribed, threatened or seduced.6 In a  word, binary voting is fickle. Let us therefore consider voting theory,  firstly to examine the workings of a binary procedure (and multi-option  voting will be discussed later on). 

Consider the following hypothesis: a committee of three persons  debating the topic of tax rates for the rich, currently assumed to be 40%.  All three are agreed, that is far too low. Ms i suggests 70% and moves a  motion to that effect; Mr j thinks 60% would be more appropriate and  proposes an amendment; while opting for a more modest 50%, Ms k moves a second amendment. Let us also assume that, as shown in Table I,  Ms i has a perfectly logical set of single-peaked preferences: 70-60-50-40;  that Mr j's preferences are also single-peaked, 60-50-40-70; while for  reasons of gender solidarity perhaps, Ms k's 3rd and 4th preferences are  

5 The term ‘all-party’ implies all the main parties. Obviously, not all 19 parties  can be represented in a Cabinet of 16! 

6 For a fuller list, see http://www.deborda.org/won-by-one/

40 Munich Social Science Review, New Series, vol. 6, 2023 

inter-changed: 50-40-70-60. After all, as the old English saying goes,  “there's nought as queer as folks.” 

Table I: The Committee's Preferences 

Preferences	Ms i	Mr j	Ms k
1st	70	60	50
2nd	60	50	40
3rd	50	40	70
4th	40	70	60

According to laid down procedures,7the committee first debates and  then chooses its more preferred amendment: 60 or 50; and by a majority of  2:1, they choose 60. Next, they decide either to adopt this amendment or  to stick with the original motion, another dichotomy, now of 60 or 70, and  both Ms i and Ms k prefer 70. Finally, it's yet another binary choice  between this substantive as it is called, 70%, and the status quo, 40%, and  again by a 2:1 majority, they prefer 40. That is it: by a 2/3rds majority,  this committee decides that they want the very thing they had agreed they  don't want. As the table shows, however, all three voters prefer 50% to 40%. So, the outcome is wrong. 

The conclusion is stark: binary voting is not only primitive and  divisive, it is also, sometimes, inaccurate. Or, to quote Riker (1988: 65),  “however democratic simple majority decision initially appears to be, it  cannot in fact be so.” Furthermore, the very procedure is so manipulable.  For the above voters with the above preferences, if the original motion had  been for 60, with 50 and 70 as the two possible amendments, the outcome  would have been 60%. Or again, if they had started with a different status  quo, let’s say of 50%, and if the motion had moved 40 with 70 and 60 as  the two amendments, the outcome would have been 50%. In effect, binary  voting can sometimes be little more than a lottery, and this is especially  true whenever there is a binary vote paradox, as there is in Table I. 70 is  more popular than 60, which we write as 70 > 60, and in all: 

70 > 60 > 50 > 40 > 70 > …. 

In such a scenario, no matter what the outcome, there is always a  majority in favour of something else. In other words, binary voting can be  

7 Many of these rules are well over 2,000 years old, established when the only known  voting procedure was binary voting.

P. Emerson: Reviewing Dutch Democracy 41 

a cause of conflict. It also means that those who decide on the order of  voting have the power to manipulate the debate and, sometimes, to  determine the outcome.8 

3. The modified Borda count MBC 

Of the other methodologies mentioned,  

(a) a plurality vote might identify the option with only the largest minority; whereas 

(b) TRS definitely identifies a democratic majority opinion,  (c) as does AV but, as I demonstrated in my contribution to MSSR Volume 3 (Emerson 2020), the AV majority opinion might not be the  same as the TRS majority opinion. 

The next two methodologies are non-preferential: 

(d) approval voting is not best suited for the political forum, not least because it incentivises the participant to be, not consensual, but  intransigent, while 

(e) range voting is even more prone to this weakness. 

The other methodologies mentioned above are  

(f) the BC,  

(g) MBC, and  

(h) the Condorcet rule.  

The last three take all preferences cast by all voters into account, always.  The Condorcet/Copeland winner is the option which wins the most  pairings. At best, the MBC winner is the option with the highest average  preference.9 In other words, the MBC is non-majoritarian and thus worthy  of further study. It works like this: In a parliament of say ten parties, there  might be up to ten different ideas on any one problem (which, it is  assumed, is non-urgent). Accordingly, one party may move a motion – option A; rather than just oppose, other parties, as if in a German  constructive vote of confidence, may propose something different – options B, C etc. If the difference is just a small amendment to one minor  

8 As was the case in Brexit. Boris Johnson ensured that ‘his deal’ was approved,  only by pitting it against that which he knew was the most unpopular of all options  – ‘no deal’. 

9In many profiles, the Condorcet winner will be the same as the MBC social  choice.

42 Munich Social Science Review, New Series, vol. 6, 2023 

paragraph, or if it is a complete re-draft, either way, every party proposing  an option must make it a complete package.10 Every option ‘on the table’  may be summarised on a computer screen and listed on a dedicated  website, with highlights showing whatever differences there might be. 

In debate, options may be tweaked, amended, composited or even  deleted, but only if the original movers agree to such changes.  Accordingly, the number of options in question may vary. If in debate this  number is reduced to a singleton, this final option may be regarded as the  (verbal) consensus.  

If not, the Speaker/Voorzitter may draw up a (short) list, usually of four  to six options. When all parties agree that their particular option (still ‘on  the table’) has been considered – that it has been included verbatim or as  amended or in composite – the members of the Tweede Kamer may cast  their preferences. If the winning option's average preference score passes  a pre-determined threshold, it may be seen as the best possible  compromise; if higher still, it may be regarded as the consensus.11 If not,  if nothing passes even the minimum threshold, then it must be assumed  that there is no agreement and that the debate should be resumed, with a  focus perhaps on those options which received the slightly higher scores. 

2.1 The Mathematics of the MBC 

The MBC should not be (but often is) confused with the BC. As I  mentioned in MSSR Volume 3 (Emerson 2020), the original formula of  Jean-Charles de Borda (though not expressed in this way) stipulated that in  any vote on n options, the voter may cast m preferences, such that 

n ≥ m ≥ 1 

and that points shall be awarded to (1st, 2nd … last) preferences cast,  according to the rule  

(m, m-1 … 1). 

rule i 

Accordingly,  

▪ he who casts only one preference gets his favourite just 1 point, ▪ she who casts two preferences gets her favourite 2 points (and her  2nd preference 1 point), 

and so on; so 

▪ those who cast all n preferences get n points for their favourite (n-1 for their 2nd choice, n-2 for their 3rd, etc.). 

10 In consensus politics, there’s no such thing as ‘a wrecking amendment’. 11 In other words, a consensus can be achieved, either entirely verbally or in a  debate which culminates in a preferential vote.

P. Emerson: Reviewing Dutch Democracy 43 

Unfortunately, even in Jean-Charles’ lifetime, the above rule i was  changed to 

(n, n-1 … 1) 

rule ii 

or 

(n-1, n-2 ...0) 

rule iii 

which, incentivises the voter to truncate his/her vote, as can happen in  approval voting. The latter formulae, rules ii and iii, are unfortunately  called a Borda count, BC… and poor old Jean-Charles is a’rolling in his  grave. 

With the MBC, however, the MPs are encouraged by the very  mathematics of the count to cast many if not a full list of n preferences. If,  then, any Tweede Kamerlid finds the policies of the Partij voor de Vrijheid PVV proposal, for example, unpalatable, they can of course submit a list of just (n-1) preferences.12 In regard to the other options, however, they  thus as it were acknowledge the validity of their parliamentary colleagues'  aspirations. So, the MBC enables the parliamentarians to be much more  inclusive than is possible in any majoritarian milieu. 

It should perhaps be noted that if the voters’ profile shown in Table I is  analysed by MBC, the committee’s social choice supports the 50% option  on a score of 9, and the social ranking is 50-60-70-40%, with scores of 9- 8-7-6; what’s more, this collective will is single-peaked. In other words,  with an MBC, they confirm their dislike of 40% and agree to increase the  tax rate to 50%. “It seems clear that Borda’s criterion is the soundest 

method of identifying the [option which] is most generally popular… or at  least the most acceptable” (Dummett 1984: 71). 

3. Power-sharing 

Now because the Netherlands has a pretty good PR electoral system,  elections tend not to lead to any one party having a majority of the seats.  But because the Parliament still uses a pretty bad decision-making system,  the above majority vote, there is then the problem of how to form a  government, which many people continue to think has to be based on a  majority coalition. 

12 And the same applies to any policies from the other extremist parties like the  Forum voor Democratie FvD and Juiste Antwoord 21 JA21

44 Munich Social Science Review, New Series, vol. 6, 2023 

The present situation is as follows: the Dutch (and many another)  democratic process consists of an open and transparent election followed  by a closed and opaque procedure of increasing duration13 as the various  parties negotiate, albeit under the guidance of a formateur, and eventually  succeed in concocting a majority coalition.14 Sometimes, its majority is  small, implying that the government is unstable – indeed, in 2017, it had a  majority of only one; and sometimes its make-up is not – how shall we  say? – wholesome. In 2002, the Christen-Democratisch Appèl CDA (on 43 seats) and the VVD (24) (People's Party of Freedom and Democracy)  joined up with the extremist Lijst Pim Fortuyn LPF (26), whose leader had  been assassinated just days before the election; these were turbulent times,  and the arrangement lasted for only five months. In 2010, VVD (31)  could have joined forces with the next largest party, the Partij van de  Arbeid Labour Party, PvdA (30); but no, and instead it coalesced with the  CDA (21) and the PVV (24) in an arrangement, gedoogakkoord, in which  admittedly the PVV did not hold any ministerial seats.15 

If, however, the democratic norm of binary voting were to be replaced  by the MBC, there would be little further justification for a majority  government (and even less for a minority administration). Instead, there  could be that which many countries advocate(d) for conflict zones like  Bosnia and Afghanistan: all-party power-sharing.  

Now most of the jurisdictions which have managed to devise some  form of inclusivity have used a formula – the Belfast Agreement relies on  a d'Hondt form of 'cherry picking', the Taif Accord in Lebanon shares out  the top positions according to the main confessional beliefs, and the  

13 In 2010, the Dutch Parliament took 127 days to form a government; in 2015, only  52, but in 2017, it was 225 days... still well short of the 2010/11 and 2019/20 world  records set by Belgium of 541 and 494 days. In all instances, forming a government involved a huge loss of time and political energy. More seriously, in 2010 and with  lives at stake, Iraq took 249 days. 

14 Italy and Greece also enjoy PR, but they avoid this problem of government  formation by awarding the winning party a large number of bonus seats… which  means, of course, that what had been a proportional electoral system can no longer  be so described. 

15 Other countries also enter into unwholesome coalitions. When Austria’s Freedom Party entered into government in 2001, the EU actually imposed  sanctions… for six months only. Both of the two main parties in Britain have  occasionally sought the support of the tiny and equally unwholesome Democratic  Unionist Party DUP, (and the fact that both of the largest British parties had a  vested interest in having a small rump of easily wooable MPs was one of many  factors which for far too long hindered any lasting solution to the NI Troubles).  Israel has also seen the unwholesome Jewish Home act as the tiny extremist tail  which wags an otherwise not quite so unwholesome dog.

P. Emerson: Reviewing Dutch Democracy 45 

Dayton Agreement for Bosnia stipulates a three-person presidency – but  all three arrangements perpetuate the very sectarianism they were  supposed to obviate. The magic Zauberformel used in Switzerland would  do the same.16 No country has yet used the obvious tactic of an election.  This is because most electoral systems enable members to choose only  ‘who’ should be in cabinet, but not ‘what', in other words, ‘who’ should be  the minister of ‘which’ portfolio. 

Hence the matrix vote17 (Emerson 2007: 61-85). Consider a very  simple example of a parliament of ten MPs electing a government of six  ministers, a Prime Minister and five others, Ministers A to E.  Accordingly, every member of this parliament would receive a ballot  paper, as shown in Table II. 

Table II: A Matrix Vote Ballot 

Preferences		The Cabinet
		PM	Minister A	Minister B	Minister C	Minister D	Minister E
1st
2nd
3rd
4th
5th
6th

Every MP of every party could then participate in selecting the  Government. If such were the case – (and this text now returns to the  present tense) – each MP chooses, in order of preference, six other  members of parliament, and inserts these names in the shaded part of the  ballot. Then, in the unshaded matrix, each MP writes six P1s, (for reasons  which will be explained later), to indicate who of these six they wish to  

16 The Swiss all-party seven-person Federal Council, whose members represent the  top four or five parties in Parliament, is based on the magic Zauberformel ratio of  2:2:2:1 or now 2:2:1:1:1.  

17 The matrix vote was first put to the test in a cross-community conference in  Belfast in 1986. Since then, it has been developed into its present format. A recent  demonstration with electronic voting was held in 2016, under the auspices of The  Irish Times in co-operation with the de Borda Institute and others. See  

http://www.deborda.org/home/2016/4/25/2016-5-ireland-let-the-dail-elect-a govt.html

46 Munich Social Science Review, New Series, vol. 6, 2023 

serve in which ministry: one P1 in each column, and one P1 in each row.  An example of a completed ballot is shown in Table III. 

Table III: A Completed Matrix Vote Ballot 

Preferences		The Cabinet
		PM	Minister A	Minister B	Minister C	Minister D	Minister E
1st	Ms g	P1
2nd	Ms l			P1
3rd	Mr p						P1
4th	Mr d					P1
5th	Mr m		P1
6th	Ms t				P1

The analysis is done in two stages, but first, a clarification: a number in  the matrix is called a sum, while any addition of sums is called a score.  The first stage of the count is a PR analysis of the data shown in the  shaded column. This could be done under the rules of any proportional and  preferential system, but the author's recommendation is for the quota  Borda system QBS (Emerson 2007: 39-60) which, because of its MBC  element, encourages every MP to submit a full ballot; and because the  matrix vote is based on a QBS proportional ballot, it thus encourages them  to cross the party divide.18  

The second stage of the matrix vote analysis is an MBC of the data in  the matrix. Every 1st preference (of a full ballot) gets 6 points, every 2nd gets 5, and so on. And the six most popular members of parliament are  then appointed, in descending order of all the sums, to the ministry for  which they get the most points. 

Let us consider the motivations of a party with, say 40% of the seats in  this parliament, which can thus expect two or at most three of its members  to be elected to cabinet. To maximise its potential, it will nominate two, at  

18 As in PR-STV, so too in QBS, parties are encouraged (by the mathematics of the  count) to nominate only as many candidates as they think they can get elected. In  a 30:30:30 constituency in Bosnia, for example, no one party would want to nominate more than two candidates, and no one sectarian group would want to  nominate more than three. If then the member of parliament were to submit a full  ballot, he/she had best cross at least the party divide if not indeed the ethno religious chasm. 

P. Emerson: Reviewing Dutch Democracy 47 most three candidates, each for a specific ministerial post. At the same  time, it must be emphasised that, to get the most points for its own  candidates, it is to every member’s advantage to submit a full and  therefore cross-party ballot – an essential pre-requisite, it is argued, of any  power-sharing polity. 

Needless to say, quite a few parties might wish to have one of their  members become PM, and doubtless too there will be keen competition  for some of the more leading portfolios, the Ministerie van Financiën for  example. Hence the P1, because members may also cast a P2 and a P3 – the letter P stands for priority – so if their particular P1 candidate does not  get appointed to be PM, the points that candidate received will be  transferred to the column(s) of his/her P2s. 

In the unlikely event of a tie,19 the decision goes to the more popular  MP (as measured in the QBS election); and if there's still a tie, it goes to  the more contended ministry (the ministry which gets the higher MBC  score, as measured in the bottom row of Table IV). 

Table IV: A Matrix Vote Count 

Preferences		The Cabinet						MBC   P1   scores
		PM	Minister  A	Minister  B	Minister  C	Minister  D	Minister  E
1st =	Ms r	20			20		10	50
1st =	Mr s	10		20		10	10	50
1st =	Ms h	30	20					50
4th =	Mr u			10			10	20
4th =	Mr m		(10)	20		(10)		20
6th =	Ms l					10		10
6th =	Ms q			10				10
MBC P1 scores		60	20	60	20	20	30	210

Let it be assumed that a parliament of ten MPs is electing a cabinet of  six ministers, and that all ten cast full ballots of six preferences. So, the  total number of points cast is 10 x (6+5+4+3+2+1) = 210. Let it also be  assumed that the count gives the results shown in Table IV: this suggests  that the first five members in the shaded column have definitely been  

19 With 150 Tweede Kamerleden casting 21 points each, the chances of any two  sums being the same are minimal; the matrix vote, however, is robust. 

48 Munich Social Science Review, New Series, vol. 6, 2023 

elected to the Cabinet, while there’s a tie for the 6th member – Ms l and Ms  q both on a score of 10 – while three members of the ten-member  parliament got no P1s at all. 

As noted above, ministries are awarded in descending order. The top  sum in the matrix is 30, (Ms h/PM), so sure enough, Ms h becomes PM.  At which point, any P2s cast for Ms r and Mr s shall be transferred to their  P2 columns but, for the sake of simplicity, it is here assumed there are  none. 

The next highest sum is 20, of which there are five, two shown in  reverse and three in tint: one is for Ms h to be Minister A, but she has  already been appointed as PM, so the (Ms h/A) 20 may be ignored; in like  manner, the (Ms r/PM) 20 may also now be ignored. The other 20s – (Ms  r/C, Mr s/B and Mr m/B, all in tint in Table IV) – show no contention for  Ms r as Minister of C; there is a tie, however, for the B Ministry; but Mr s is more popular than Mr m, so Mr s wins. Whereupon Mr m’s 20 P1s are  transferred according to their P2s, let us say 10 to each of A and D – shown in Table IV in brackets. – The situation, so far, is shown in Table  V. 

Table V: A Matrix Vote Count, in Progress 

Preferences		The Cabinet				MBC P1 scores
		PM	Minister   A	Minister   B	Minister   C	Minister   D	Minister   E
1st =	Ms r				20			50
1st =	Mr s			20				50
1st =	Ms h	30						50
4th =	Mr u						10	20
4th =	Mr m		(10)			(10)		20
6th =	Ms l					10		10
6th =	Ms q			10				10
MBC P1 scores		60	20	60	20	20	30	210

There is now another tie, with four candidates – Messrs u, m, l and q – all  competing for the last three appointments, the Ministries A, D and E, each  on a person/post sum of 10 points. With scores of 20 (as shown in the  right-hand column), Messrs u and m are the more popular candidates, so  Mr u is appointed to the E ministry; and given that the A ministry is  uncontested, whereas D has another aspirant, Mr m takes on the Ministry  of A. This leaves Ms l and Ms q, both on a score of 10, but only Ms l has  any points for D, so she fills the final post.

P. Emerson: Reviewing Dutch Democracy 49 

All things being equal – that is, if the election takes place without  threats or bribes and the like – the outcome is bound to be a Cabinet in  which, individually, every minister is well suited to his/her appointed  ministry, while collectively, the Cabinet then represents the various parties in parliament, each (approximately) in their proportional due. The result of  the above hypothetical vote is shown in Table VI, and please note, the  order of ministries has been re-arranged in order, left to right, of their P1  scores, a measure of the degree of contention of each ministry. 

Table VI: The Elected Cabinet 

Preferences		The Cabinet						MBC P1   scores
		PM	Minister  B	Minister   E	Minister   C	Minister   D	Minister   A
1st =	Ms h	30						50
1st =	Ms r				20			50
1st =	Mr s		20					50
4th =	Mr u			10				20
4th =	Mr m						10	20
6th	Ms l					10		10
Others								10
MBC P1 scores		60	60	30	20	20	20	210

To conclude, the Netherlands’ entire democratic electoral process would  therefore consist of one open and transparent election, in which the people  choose their parliamentarians; followed about a week later by another  equally open and transparent election, in which the latter choose their  government. 

3.1Will it Work? 

It should first be recognised that binary majority rule does not work well,  anywhere. In the USA, it often makes governance dysfunctional. While  in conflict zones, not only in Northern Ireland but in the Balkans, the  Caucasus, the Ukraine20 and not least in the Middle East, it is often (not  the perhaps but) a cause of that conflict.  

20 When Ukraine became independent in 1991, the EU (or EC) supported a polity  based on majority rule. In 2014, however, when the protests in Maidan became  horribly violent, the EU changed its mind in favour of power-sharing, and a  delegation rushed over to Kiev… too late; it arrived on the very day Viktor 

50 Munich Social Science Review, New Series, vol. 6, 2023 

Table VII: Potential Governments of National Unity 

Country	Last   election	Party	Number  of seats in   parliament	Expected   number   of   cabinet   seats
Problematic
Austria	2019	FPÖ, Freiheitliche Partei Österreichs	31/183	3/15
Germany	2017	AfD, Alternative für Deutschland	94/709	2/16
Ireland	2020	SF, Sinn Féin	37/159	3/15
Netherlands	2017	PVV, Partij voor de Vrijheid	20/150	2/16
UK	2020	DUP, Democratic Unionist Party	10/650	0/25
Purposeful
Israel	2020	Arab List	15/120	7/37
Turkey	2018	HDP, Halkların Demokratik Partisi	67/600	2/19
Problematic
Israel	2022	Religious Zionist	14/120	4/31
		Shas	11/120	3/31
		United Torah	7/120	2/31
		Otzma Yehudit	6/120	2/31
		Noam	1/120	-/31

That said, as shown in Table VII, GNUs could have quite a few  consequences: in the first four countries, it could mean that extremist  parties were ‘at the table’ but – as indicated in the final column – only in  their proportional due. In the two ‘Purposeful’ rows, such a GNU could  be the very basis of a long overdue more inclusive polity, a cause for hope  for both countries’ Arab and Kurd minorities. 

In the Netherlands, (as too in Germany and Ireland), many persons in  other political parties have often said that they would not, indeed could  not, sit in cabinet with those of the PVV (as too AfD in Berlin or Sinn Féin  in Dublin). In a consensual polity, however, there is little to fear: no  faction (PVV) of only two persons can dominate a Cabinet of 16,  especially if the latter is working in consensus. What’s more, most elected  politicians want to be re-elected. To do that, they must be effective. And  to do that in a consensual milieu, they must moderate their extremism. At  the same time, if governance is to be subject to just a majority coalition,  there is always the danger of an extremist administration in which the dog  

_______________________ 

Yanukovych ran into exile.

P. Emerson: Reviewing Dutch Democracy 51 

is wagged by at least one of many tails … as is currently the case in Israel.  I rest my case.  

Meanwhile, in other countries, such a robust and proportional  methodology could well facilitate that which in many instances is an  otherwise intractable problem, as was the case for example in Afghanistan,  where the very western concept of political party is a little more – let us  say – flexible. Generally speaking, however, when needs must, parties of  supposedly diametrically opposite views have often shown themselves to  be able to work together, sometimes in grand coalitions, or occasionally,  as in war time, in all-party coalitions. The Covid and Climate Change  crises are no less existential (Emerson 2022: 86f).  

Politics is indeed the art of the possible. Consensus voting – the MBC  in decision-making, QBS in elections and the matrix vote in governance – is its science. 

4. Conclusion 

It is quite extraordinary, but many countries have an electoral system which allows the voter to cast only one preference. It is as if he who  supports the CDA, for example, regards all the other parties, from PvdA to  PVV, with equal disdain. This is obviously not the case. In other words,  if she is allowed to cast only one preference, the voter cannot express her  opinion accurately. The overall election result, the collation of all this  inaccurate information, will therefore be imprecise as well – (and this is  especially true in a single-preference system like the UK’s first-past-the 

post). A preferential procedure like (PR-STV or) QBS could help to  overcome this problem (rather more readily, it is suggested, than a Tsebelis-Crosson multi-vote procedure). 

What is even more extraordinary is the fact that so few question  majority voting. As I wrote in MSSR (Emerson 2020), binary voting is  Orwellian: ‘this’ good, ‘that’ bad. Yet it is ubiquitous. It is even enshrined  in the North Korean Constitution. (Not that it is often used in  Pyongyang.)21 One can understand why politicians like majority voting: it  invariably allows them to write the question on the referendum or  

21 Socialist Constitution of DPRK, Article 97, para 2: “…decisions of the Supreme  People’s Assembly are adopted when more than half of the deputies attending  signify approval by a show of hands.” In like manner, Para 3 stipulates a two thirds majority for any constitutional amendments. The North Korean parliament  meets only once or twice a year! (Article 92).

52 Munich Social Science Review, New Series, vol. 6, 2023 

parliamentary ballot, and that question is usually the answer. But why do  members of the media and academia, along with the founding fathers of  the US Constitution in 1776, as well as those involved in drafting the  Grundgesetz in the 1949 Parliamentary Council in Bonn, and those in  Dayton in 1995 writing the Bosnian Peace Accords, and those in Belfast  drawing up the 1998 Good Friday Agreement, and those concerned with  the 2001-3 Convention on the Future of Europe … why do most of these  good people just ignore multi-option voting procedures? They sometimes  tweak majority voting, and hence consociational voting (in Belgium,  Bosnia, Cyprus and Northern Ireland), along with qualified majority  voting (in the EU) and constructive votes of confidence (in Germany).  But preferential voting? The ideas of Ramón Llull, Nicholas Cusanus,  Jean-Charles de Borda, Rev. Charles Dodgson (Lewis Carroll) and so on,  apparently, count for little. So, problems are still reduced to dichotomies,  or series of dichotomies. So, politics continues to be adversarial. So, minorities still argue with majorities. Often they fight. And the world  struggles on, from one blunder to the next.  

It is strongly suggested that the need for the Netherlands to adopt a  more inclusive polity should be seen in this international context. 

References 

Brouwer, P., and K. Staal (2020), “The Future Viability of the Dutch  Democracy; A Model Case,” in: M.J. Holler (ed.), Improving  Democracy (I) and Robots, Munich Social Science Review, Volume 3:  7-53. 

Dummett, M. (1984), Principles of Electoral Reform, Oxford: Oxford  University Press. 

Emerson, P. (2007), Designing an All-Inclusive Democracy, Heidelberg:  Springer. 

Emerson, P. (2020), “Can Rights BeWrong? Towards a Less Majoritarian More Inclusive Democracy,” in: M.J. Holler (ed.),  Improving Democracy (I) and Robots, Munich Social Science Review, Volume 3: 93-111. 

Emerson, P. (2022), The Punters’ Guide to Democracy, Cham: Springer Nature. 

Nurmi, H. (2020), “Remarks on ‘The Future Viability of Dutch Demo cracy’,” in: M.J. Holler (ed.), Improving Democracy (I) and Robots,  Munich Social Science Review, Volume 3: 55-62.

P. Emerson: Reviewing Dutch Democracy 53 Riker, W.H. (1988), Liberalism against Populism, Long Grove: Wave land Press.