Electronic Journal of Comparative Law, vol. 11.1 (May 2007),

Constitutional Referenda in the Netherlands: A Debate in the Margin

L.F.M. Besselink

Readers are reminded that this work is protected by copyright. While they are free to use the ideas expressed in it, they may not copy, distribute or publish the work or part of it, in any form, printed, electronic or otherwise, except for reasonable quoting, clearly indicating the source. Readers are permitted to make copies, electronically or printed, for personal and classroom use.

The issue of constitutional referenda touches on the very foundations of the constitutional system of a country. It inquires into the nature of the power to make and to amend a constitution, and into how to understand this power and the procedures involved. In particular, it raises fundamental questions regarding the democratic nature of constitutional politics as well as the nature of democracy at the basis of a constitution.

This contribution on the Constitution of the Netherlands could have been very short indeed, much shorter than it is. The present Constitution of the Netherlands does not provide for a referendum, nor has it been practised in the Constitution’s making or its amendment. With one or two exceptions, the constitutional referendum is hardly recognized as being of its own kind in this country.

Yet, as we shall see, constitutional referenda have been looming large at the margins of the system. They are at the temporal horizon of the constitutional system from the very beginnings of the Kingdom until the recent rejection of the Treaty establishing A Constitution for Europe. They are at the margins of its territorial scope, as the overseas countries of the Kingdom have repeatedly consulted their population as to the desired constitutional future of the Kingdom. And they are at the horizon of political debate, or rather – with one exception – in the margin of a series of proposals throughout the 20th century to introduce a general legislative referendum.

The constitutional referendum is also looming at the margins of the present method of constitutional amendment in a quite different manner: although the Constitution does not provide for a referendum or anything equivalent to that, paradoxically the present method of amending the Constitution is taken by most constitutional lawyers to involve a consultation of the electorate on pending amendments. We hold this view for erroneous on several scores. This kind of misunderstanding as to the precise nature of the constitutional process is in itself significant for the type of constitution we are dealing with.

We shall discuss these various elements in the present contribution. After a brief general characterization of the constitutional system, we will first discuss the present method of constitution making. Next we will have a brief glance at the use of constitutional referenda in the constitutional history of the Netherlands, at the use of consultative constitutional referenda in the overseas countries, and finally at the various proposals for introducing constitutional referenda. We will round this essay off with a number of concluding reflections.

1.The Nature and Character of the Constitution of the Netherlands

When we distinguish between revolutionary constitutions and historical, incremental constitutions, the constitution of the Netherlands is definitely of the latter type. There is a single document called Grondwet. In 1815, due to the union of Belgium to the Northern Netherlands, it had both a Dutch and French official version. In French it was, significantly, not called Constitutionbut Loi fondamentale. This gives a better rendering than the current English translation which we shall use here as well, which translates Grondwetas Constitution of the Netherlands.

This Grondwet does not encompass the whole of the constitution, nor does it contain the most important constitutional rules. Of higher rank than the Grondwetis the Statuut voor het Koninkrijk der Nederlanden, Charter for the Kingdom of the Netherlands, which concerns the relations between the European country and the two overseas countries, the Netherlands Antilles and Aruba. Not all fundamental constitutional rules can be found in either of these documents. The fundamental rule of the governmental system – the rule of no-confidence which makes it a parliamentary system of government – has remained a rule of unwritten, customary constitutional law to this day, just as ministerial responsibility has been expressed in the Constitution only in a laconic and incomplete manner.[1]

Apart from other rules of constitutional law contained in sources outside the Constitution [Grondwet], there is a fair amount of consensus that the role of the Constitution is primarily to codify rules and principles which have become generally accepted, rather than to modify public society and to dominate and steer the political system.

Most of the great constitutional modifications took place in the 19th century and ended with the democratic reforms of the early 20th century. These modifications basically concerned the governmental system, which changed from autocratic monarchical rule – albeit under a constitution – towards a system of quasi-aristocratic parliamentary government in 1848 as confirmed by the parliamentary practice of the 1860s; and finally it were constitutional provisions which transformed the system from aristocratic towards democratic as a result of debates raging since the 1870s and finding an end with the revision of 1917 and its final confirmation in the constitutional amendments of 1922.

All these modifications and reforms took place by incremental changes to the Constitution of 1814/1815. The changes required no new republics, not even new constitutions. Although there is some disagreement on whether the original of the present Constitution dates back to 1814 or 1815 – which in itself is significant – we are still constitutionally governed by that 1814/1815 Constitution as amended since. Continuity is the hallmark of the constitutional system, notwithstanding intermittent political upheaval, which, on occasion, led to constitutional accommodation; a real political revolution is hard to trace in Dutch constitutional history since 1848. And even 1848 was not accompanied by truly revolutionary movements in politics – the liberal movement in the Netherlands was a movement of reform, not a revolutionary movement.

As we shall see, the fact that in the Netherlands the Constitution does not occupy that exclusive and elevated position which it does in other continental European countries has also meant that the constitutional culture is relatively weak. Constitutional debate is nearly absent and, if engaged in, is not very profound nor characterized by intellectual rigour, but dominated by considerations of political convenience. It is not a hazardous guess to hold that politicians are hardly aware of constitutional rules and principles beyond one or two of these. The Constitution is an instrument which in the end is not very different from any other piece of legislation. This sense is perhaps reinforced by the fact that acts of parliament cannot be reviewed by courts for their compatibility with the Constitution (though they can be reviewed against self-executing treaty provisions such as those which are found in, for instance, human rights treaties).

This all has its effect on issues regarding constitutional referenda. But before we can say anything about these, we must indicate the present method for amending the Constitution, because it is within that method that it might have a role to play.

2.The Present Role of the Electorate in Constitution Making

2.1.Pouvoir Constituante

The background we have just sketched, may help understand various traits of the constitutional law of the Netherlands. One of these is that in the literature there is little concern to identify the pouvoir constituante, neither historically, nor in terms of present-day constitutional theory. There is no clear concern to identify who has held or holds original power to constitute the state, its institutions and the principles and rules on which they are to operate. Moreover, sovereignty is considered a concept which is not helpful in making a constitution like the Dutch work, nor for understanding its operation. As a somewhat singular exception to continental Europe’s constitutional doctrine, sovereignty has no constructive role to play in the treatises and textbooks on constitutional law of the Netherlands.[2] This approach has deep historical roots, which go back to the more than two centuries of the Republic of the United Provinces which did not claim sovereignty over the provinces it united, and in which no institution of this overarching Republic claimed sovereignty either.[3] During the 19th century that other concept of sovereignty – sovereignty of the people – was too divisive a principle to be invoked successfully as a founding or legitimating concept.

An authentic, original constitution making power, a true unconstituted pouvoir constituante, existing so to say extra ordinem, cannot be identified in the Netherlands constitutional history. The Constitution is much rather a product of the vagaries of history, consolidated by incrementally taking up what has emerged as a new or not so new consensus.

As a consequence of this state of affairs the literature speaks rather loosely about the grondwetgever, constitution-making power, without any clear distinction between the power to make a new Constitution and the power to amend it within the bounds of the Constitution’s own provisions on constitutional amendment – and so shall we in this essay. Nevertheless, at times the Constitution has been amended in an unconstitutional manner. This was the case at any rate in 1815[4] and 1840,[5] and, as we shall see below, at least arguably also the recent amendment of 2005 was unconstitutional. But none of the subsequent constitutions have been considered revolutionary.

Also institutionally, this constitution-making (or rather: constitution-amending) power is hard to define. It requires a careful analysis of the provisions on which the process of constitutional amendment is based, as well as of constitutional practice. We shall undertake this analysis with a view to ascertaining the manner in which the electorate cooperates in the exercise of the constitution-making power. We shall conclude that this electorate (the ‘people’ as one would have it in certain constitutional systems) does not have constitution-making power, neither in fact nor in law.

2.2.Constitutional Provisions on the Procedure for Constitutional Amendment

Since 1814, the Constitution has been amended 22 times, usually in packages of amendments comprising the change of one or more provisions: 1815, 1840, 1848, 1884, 1887, 1917, 1922, 1938, 1946, 1948, 1953, 1956, 1963, 1972, 1983, 1987, 1995, 1999, 2000, 2002, 2005. The process of amending the Constitution is determined in Article 137 of the Constitution.[6]

The essence of the procedure is that the amendment has to be passed twice in parliament and be ratified twice by the government; these are called the two ‘readings’. Until 1996, we could say more simply that an amendment had to be adopted by two successive parliaments and ratified by two successive governments. Since 1996, however, no new elections for the Upper House are necessary for it to adopt an amendment. So two different Lower Houses deal with the amendment, but in principle it is the same Upper House deals with it twice.

The ‘first reading’ takes the form of a normal Act of Parliament ‘which states that an amendment to the Constitution as proposed by the Act shall be considered’ (Art. 137, first paragraph, of the Constitution). This Act requires the normal majority of the votes in each of the two Houses of Parliament.

After publication of this Act, the Lower House is dissolved (Art. 137, third paragraph, Constitution) – a rule introduced in 1848.

After the election of the new Lower House, each of the Houses of Parliament ‘shall consider, at second reading, the proposed amendment […]. They can only pass it with at least two thirds of the votes’ (Art. 137, fourth paragraph, Constitution). This is the ‘second reading’.

After governmental ratification, the amendment enters into force immediately upon publication.

In constitutional practice, the dissolution prescribed by Article 137, third paragraph, of the Constitution, is timed in such a manner that the elections coincide with the periodic general election of the Lower House. As we said, the constitutional amendment of 1996 scrapped the dissolution of the Upper House, This was justified as follows. The Upper House is elected by the Provincial Councils immediately after the direct election of the latter. As the dissolution of the Upper House is not accompanied by the dissolution of the Provincial Councils, in practice the Provincial Councils would tend to re-elect the same Upper House.[7] This was considered to be an ineffective ritual.

2.3.The Nature and Meaning of the Dissolution of the Lower House

The predominant view in present-day constitutional literature is that the dissolution of the Lower House, as part of the process of constitutional amendment, is a consultation of the electorate on the amendment. Often this is taken so far as stating that the newly elected parliament is to decide in accordance with the wishes on the amendment expressed by the electorate. This view goes back to the writings of Thorbecke in the first half of the 19th century.[8] This view is, at least under present circumstances, erroneous both in fact and, it is submitted, also in law for the following reasons.

It is erroneous in fact because the electorate does not pronounce itself on the amendment at all. It is true that the amendment may theoretically be a prime issue during the elections. However, the only time that this may be said to have been the case was in 1917, when the general franchise was introduced as well as equal financial treatment of public and denominational schools. The amendment of 1948, which created the (ultimately abortive) possibility of a kind of commonwealth confederation of the Netherlands and Indonesia (as well as Surinam and the Netherlands Antilles in the West), was another occasion where the politics of decolonisation at the basis of the constitutional amendment played a role in the elections. But in none of the other cases during the 20th century amendments and the 21st century did the amendment play any role during election campaigns. So as a matter of fact, dissolutions are not aimed at consulting the opinion of the electorate on the amendment pending.

This conclusion is reinforced by the following arguments. Since the introduction of an electoral system of proportional representation in 1917, elections have never produced clear majorities but only minorities. Even if a constitutional amendment were to be limited to one issue only – which it has rarely been – then it would be a rare event to have any clear position with clear results. The fact that parliament after the dissolution continues also as a regular parliament, complicates this further. It means that the issue of the constitutional amendment cannot reasonably be the one and only consideration in casting a vote.

Until 1995, the Upper House was dissolved simultaneously with the Lower House except, in 1995, the dissolution of the Upper House was postponed to coincide with the periodic election of the Upper House by the (newly elected) Provincial Councils. In 1996, as we mentioned, the dissolution of the Upper House with a view to deciding on a pending constitutional amendment was abolished altogether. The consequence of this is that one of the two Houses of Parliament with decisive powers on the amendment of the Constitution is necessarily provided not merely with an older democratic mandate, but with one which is necessarily unrelated to any particular constitutional amendment for the simple reason that at the time of the elections of the Provincial Councils, which subsequently elect the Upper House, there is (usually) no constitutional amendment yet passed at first reading; so the amendment cannot possibly play a role at these elections. Yet, this House has an identical power to adopt or to reject the amendment of the Constitution as the Lower House, which was renewed for the purpose. Under this circumstance, it is incoherent to conceive of the elections as a kind of plebiscite which translates into a newly composed parliament deciding in conformity with the wishes expressed by the electorate on the amendment: if there would be any plebiscitarian aspect to the dissolution (which there is not) this could, since 1996, only be the case for one of the two Houses with constitution-making powers.[9] If the aim of the dissolution of the one House would be to give the electorate the chance to approve or disapprove of the amendment, this approval or disapproval cannot possibly bind the other half of parliament (the Upper House) which is not dependent on any new election.

It is not the newly elected Lower House which alone has decisive constitution-making power; it shares it with the old Upper House whose mandate has not been renewed to adapt it to its new role of makers of the Constitution. The Upper House is turned from a law-maker into a constitution-maker, as if by chance, through the coincidence of the adoption of an Act at first reading and the dissolution and the forming of a new Lower House with constitution-making powers. It certainly is not turned from a law-maker into a constitution-maker by any decision of an electorate, not even by its own electorate, being the (themselves directly elected) members of the Provincial Councils. As there is no trace of any direct or indirect decision of an electorate that could possibly be at the basis of the subsequent decision of the Upper House to approve or not to approve the amendment (disapproval requiring only one third of the vote (minus one), approval two thirds of the vote), it has become impossible to say that at the basis of parliament’s decision there is some quasi-plebiscitarian consultation of the electorate.