• 1 Introduction

  Political scientists often use Belgium as an ideal case study for discussing processes of government formation and of caretaker cabinets. Combined with its complex multilevel institutional architecture and its enduring regionalist tensions, these processes have attracted much attention from the international community. The various episodes of the lengthy federal government formation even kept the international media in suspense over the last decade. The fact that Belgium had a caretaker government throughout its successful EU presidency term in 2010 impressed many European observers.
  Caretaker periods mark the transition between the termination of one government and the start of another. If the end of a cabinet and the kick-off of a new one can be considered as gold mines for political scientists working on elections, executives and ministerial careers, the inter-bellum period did not receive the same research attention. Literature on caretaker cabinets, in particular, is scarce (Boston, Levine, McLeay, Roberts & Schmidt, 1998; Courtenay Ryals & Golder, 2010; Schleiter & Belu, 2015). Yet several scholarly works on caretaker governments in Belgium have been published in prestigious journals, and it even constituted the core topic of an entire debate section of European Political Science back in 2012. These studies often focused on the 2010-2011 political stalemate and limited themselves to two main phenomena: (1) how can we explain these long periods of caretaker government and (2) how can we explain that the Belgian political system did not collapse – but actually did quite well – under a caretaker cabinet?
  Despite these relevant academic contributions with regard to caretaker cabinets, there are no systematic data or studies available that demonstrate the different ways in which such cabinets diverge from others (Courtenay Ryals & Golder, 2010). Also, there have been no thorough attempts to disentangle the various kinds of caretaker periods. Consequently, authors have used diverse definitions of what a caretaker government actually is. Furthermore, there is no agreement on the exact moments when a caretaker period begins and ends. Therefore, this research note has two objectives. First, it explores the definition, characteristics and how caretaker cabinets are best measured timewise. Second, it provides a detailed typology of the different caretaker periods, a tool that will subsequently allow us to compare caretaker cabinets across time and space. Our assessment is based on previous academic works and on a detailed analysis of the Belgian caretaker cabinets between 2007 and 2020.

• 2 Caretaker Cabinets: Definition and Characteristics

  In order to prevent a void in political decision making between two different governments, countries have implemented various procedural strategies aimed at bridging such a gap. In presidential regimes, for example, the newly elected president does not take up power immediately after the elections but rather enters into executive office a few weeks later. This period varies from country to country. For instance, in the USA, Donald Trump took the oath of office on 20 January 2017, i.e. no less than 73 days after the presidential elections, and 32 days after the actual vote of the electoral college. This time interval is deemed necessary to prepare for the transition between the two administrations and cabinets. On the other hand, in parliamentary regimes the new cabinet usually takes power only after its investiture vote in parliament. Yet this investiture usually takes place a few days or even weeks after the elections. This custom affects both Westminster and consensus democracies but is more frequent in countries with proportional representation (Boston et al., 1998). Moreover, there coexist subnational variations of these caretaker conventions, for example in Australia (Davis, Ling, Scales & Wilkins, 2001; Tiernan & Menzies, 2007). But, overall, it is this specific period in the lifetime of a government, located in between the former and the new cabinet, that is called a ‘caretaker’ government.¹x The concept of ‘caretaker government’ is the most widely accepted term in the literature but may take different names in different countries; for example ‘interim government’ in Iraq, ‘demissionary cabinet’ in the Netherlands, ‘government in functions’ in Spain or ‘government of current affairs’ in Belgium.
  After its resignation or its removal by parliament, a cabinet is supposed to cease its activities immediately. The reasoning is that the dismissed cabinet cannot take any further decisions or actions that would compromise the future responsibilities of the next cabinet. However, it is necessary to avoid a complete absence of the executive power, as this could be detrimental to the country. Therefore, the resigning or removed cabinet cannot immediately leave office and instead needs to remain in power until its successor is appointed (Schleiter & Belu, 2015). This is the first characteristic of a caretaker government, which ensures that there is continuity and that the country is never without a functioning executive. In that respect, some authors have described caretaker cabinets as fulfilling a ‘bridging role’ between duly mandated governments and have referred to them as ‘interim governments’ (McDonnell & Valbruzzi, 2014).
  In that sense, caretaker cabinets have to be distinguished from those cabinets that are specifically appointed to bridge the transition between two cabinets or when their sole purpose is to bring the country to (early) elections: these are considered as interim or transitory cabinets.²x Courtenay Ryals and Golder (2010) label these cabinets ‘new caretaker’ governments, while the type of cabinets studied in this note corresponds to what they call ‘continuation caretaker’ governments. Many examples of such transitory cabinets can be found in semi-presidential systems and in Central Europe (Amorim Neto & Strøm, 2006; Courtenay Ryals & Golder, 2010; Hloušek & Kopeček, 2014; Muller-Rommel, Fettelschoss & Harfst, 2004). In Sweden, the transitory cabinets are supposed to be apolitical and non-partisan and are therefore composed of technocrats (Beckman, 2007; Larsson, 1994). Non-partisan transitory governments have also been observed in Bangladesh, Bulgaria and Italy and, to a lesser extent, in Finland and Portugal (Courtenay Ryals & Golder, 2010; Hloušek & Kopeček, 2014; Magone, 2000; McDonnell & Valbruzzi, 2014; Zafarullah & Yeahia Akhter, 2000).
  A second important characteristic of a caretaker government concerns the limited scope and range of actions of the executive. According to Van Aelst and Louwerse (2014), it is a situation in which the active government can do little more than handling ‘current affairs’. The government is only ‘taking care’ of the cabinet functions and duties whose continuity seems essential. The basic principles a caretaker government adheres to are twofold. First, the cabinet refrains from taking decisions that may burden the incoming government, whereby it restricts itself to preserving the ‘policy status quo’ (Boston et al., 1998; Davis et al., 2001; Schleiter & Belu, 2015). Given that a change of government is probably imminent, it is considered inappropriate to bind the incoming government by committing to significant new political initiatives. Second, the caretaker cabinet does not undertake new political initiatives and simply postpones all significant decisions until the new government takes over. Depending on the country, the caretaker government can undertake actions within a larger or smaller range of policies.
  Several scholars have outlined additional characteristics. In their works on minority governments, Herman and Pope (1973) remark that caretaker governments have not only limited freedom of action but also a limited life span. They regard caretaker governments as ‘default administrations’ that remain in power for just a limited and prearranged period of time. Yet the issue of a time constraint that actively limits the life span of the cabinet is not present in later academic works. Moreover, this feature actually seems contradictory to the initial definition of a caretaker government. Indeed, the notion of a caretaker government rests on the idea that this government remains in power until its successor is finally appointed. In the words of Hooghe (2012), “No matter how long that might take, the earlier government simply has to soldier on.”
  In some countries, such as New Zealand or the UK, underdeveloped caretaker conventions can leave the country vulnerable to political crisis and controversy (Boston et al., 1998; Schleiter & Belu, 2015). Take, for instance, Australia where caretaker conventions are not legally binding rules (Davis et al., 2001). An additional problem is that caretaker governments do not enjoy sufficient political legitimacy. First, caretaker cabinets may have lost the vote of confidence in parliament, which means that their ministers do not enjoy the trust of the MPs anymore. Second, the cabinet parties have not yet gone through a new ballot box verdict and/or through a new vote of confidence in parliament. This lack of political control then conflicts with the core democratic principle of the political accountability of ministers vis-à-vis the legislative assembly. Indeed, since the former government has already resigned or has been removed, its actions can no longer be controlled by parliament (Baeselen, Toussaint, Pilet & Brack, 2014). Thus, the legislative power can no longer cast a vote on a motion of no confidence against an – already – removed executive. But even if parliament cannot sanction the caretaker government as such, it can still control it in principle. For instance, it is not rare to see their ministers being questioned publicly during plenary or committee meetings. Parliamentary consent is also still needed to pass any new legislation.

• 3 Measuring Caretaker Cabinets in Belgium

  Rather surprisingly, there is no formal definition of a caretaker government within the Belgian legal framework. As in many other parliamentary regimes (Boston et al., 1998), the concept of caretaker government or cabinet is not present in the constitution, even though the Belgian constitutional framework is quite robust in ensuring the continuity of the governmental function (Hooghe, 2012). In regard to budgetary issues, a complete governmental deadlock, as happens in the United States, is virtually impossible in Belgium. Caretaker governments are only briefly mentioned in the special Law on Institutional Reforms (d.d. 8 August 1980), which stipulates that “as long as it has not been replaced, the demissionary cabinet remains caretaker”.
  Thus, the definition of a caretaker government is determined by customary law and practice. These conventions are legally enforceable by the Council of State: cabinet’s administrative acts that do not respect these conventions run the risk of being annulled (Brans, Pattyn & Bouckaert, 2016). A common standard acceptance is that a caretaker government limits itself to just three main types of policies: daily matters, ongoing matters (i.e. policy continuity) and urgent matters. Hooghe (2012) adds to this a fourth set of policies that ensures the country’s stability and fulfils its international obligations. Also, the range of actions of the caretaker cabinet is sometimes further ‘specified’ by means of own public communication. This was the case during the 2010-2011 political crisis, when the services of the prime minister and the minister of Budget released a circular letter on 7 May 2010 on the future responsibilities of the caretaker government (Brans, 2012).
  While there is a consensus among scholars on the conceptual definition of a caretaker government in the context of Belgium, expert opinions are divided when it comes to the exact measurement of caretaker periods. Following the works of Courtenay Ryals and Golder (2010), we understand that the exact duration of caretaker cabinets depends heavily on its actual measurement. When does this caretaker period start, and when does it end? Is the start of the time interval located before or after a general election, or does it cover both periods? Does it also include the process of government formation, or can a caretaker government be in place independently of parallel party negotiations?
  In brief, to determine the date of commencement of a caretaker period, we observe that there are generally three events to consider: (a) the King’s acceptance of the government’s resignation (Bouckaert & Brans, 2012; Brans, 2012), (b) the sitting government loses a vote of confidence (Schleiter & Morgan-Jones, 2009), or (c) the parliament is dissolved (Davis et al., 2001; Schleiter & Morgan-Jones, 2009). There is an ongoing debate as to whether the formal announcement of new elections should be added to this list as a fourth relevant moment (Boston et al., 1998). The end date of a caretaker period is shaped by distinct events such as (a) new elections (Schleiter & Morgan-Jones, 2009), (b) the end of negotiation talks producing a new coalition government (Bouckaert & Brans, 2012), or (c) the formal appointment of new ministers by the King (Hooghe, 2012).
  In fact, a large array of phenomena and events might classify as reference points for a ‘caretaker government’. For instance, Brans et al. (2016) distinguish between two archetypes of caretaker periods: (a) between the parliament’s dissolution (or when an incumbent government loses a vote of confidence in parliament) and a general election and (b) between a general election and the formation of a new government. On the other hand, Boston et al. (1998) identified two alternative types of caretaker periods: (a) between an election and the swearing-in of a new government and (b) between a government resigning or losing a vote of no-confidence in parliament and the formation of a new government.³x These authors are less clear about the presence of a caretaker cabinet in two other possible scenarios: (a) the period between the dissolution of parliament and the following general election (provided that the government still enjoys the confidence of parliament) and (b) the period between the announcement of an election and the dissolution of parliament (see, for instance, Simms, 2011 for a discussion of variants of political leadership during this period in Australia and New Zealand). Finally, the most developed typology of caretaker periods is probably the one Baeselen et al (2014) proposed as they distinguish between four types of periods: from the cabinet resignation until the dissolution of parliament; between the dissolution of parliament and the elections; between the elections and the installation of the new parliament, and between the installation of the new parliament and the swearing-in of the new cabinet.
  These aforementioned procedural typologies diverge from the formal judicial interpretations of this uncategorised political phenomenon. This comes as no surprise, since the actual dynamic of caretaker governments is also heavily shaped by political evolutions in practice. If we apply a too narrow definition in the study of this phenomenon, we may fail to detect some valuable information about this political concept. Therefore, in this research note we call for a broader and more comprehensive understanding of this ‘moving target’. Indeed, we believe that enlarging our conceptual scope of attention will improve our abilities to effectively analyse and explain the political relevance of caretaker governments. More specifically, we consider a caretaker period as the interval during which either of the two branches (executive or legislature) does not enjoy its full powers. The exact caretaker period then possibly starts on the day of two distinct phenomena: the resignation of the cabinet or the dissolution of the parliament (if the cabinet still enjoys its full capacity). It ends on three possible occasions: the withdrawal of the resignation of the cabinet, the vote of confidence of the new cabinet or the installation of a new parliament (provided that the cabinet still enjoys its full capacity).
  In other words, we opt for a more all-encompassing definition of what a caretaker cabinet comprises, i.e. a more generous measurement in comparison with previous works. Take, for example, the famous caretaker period⁴x This period was labelled in the international media as the ‘world’s longest government formation period in modern history’. at the end of the Leterme II cabinet in 2010-2011. Some scholars calculated that it lasted for 541 days (see, for instance, Baeselen et al., 2014) between the elections on 13 June 2010 and the oath of the new cabinet in the hands of the King on 6 December 2011, whereas others found that it lasted for 589 days (see, for instance, Bouckaert & Brans, 2012) between the King’s acceptance of the cabinet resignation on 26 April 2010 and 6 December 2011. In contrast, our measurement then leads to an even longer caretaker period: 597 days, from the day of the government’s resignation (22 April 2010) up to the vote of confidence of the new cabinet in parliament (10 December 2011).
  When using this definition for the whole period under study, it means that Belgium was governed by a de facto caretaker cabinet during no less than 1,485 days between 2007 and 2020. In other words, this corresponds to more than four (!) full calendar years, or to 29.04% of the entire period between 1 January 2007 and 31 August 2020. Indeed, we argue that observing such a longer interval will allow for a more complete analysis of the political dynamics that are at play here. Table 1 presents the duration of the different Belgian coalition governments from 2003 to 2020, including the duration of the respective caretaker periods (in days).⁵x For the clarity of the argument, we included the whole period of the Verhofstadt II cabinet (2003-2007) in this table. In the remainder of the research note, we focus solely on the 2007-2020 period. For instance, of his 1,026 days as Belgian prime minister, Yves Leterme spent more than half (59.55%) as the head of a caretaker cabinet. Far behind this dubious record, Charles Michel and Elio Di Rupo spent, respectively, 17.26% and 16.73% of their prime ministership in a caretaker period. On the contrary, Herman Van Rompuy spent only three days as prime minister of a caretaker cabinet.

  Table 1 Belgian cabinets (2003-2020)

  Cabinets	Cabinet duration (in days)	Caretaker period (in days)
  Verhofstadt II	1,623	233
  Verhofstadt III	90	2
  Leterme I	285	16
  Van Rompuy	330	3
  Leterme II	741	595
  Di Rupo	1,040	174
  Michel I	1,520	5
  Michel II	322	313
  Wilmès I	142	142
  Wilmès II	Ongoing	2
  Total		1,485

  Note: Our analysis ends in August 2020.

  Timeline of the 2010-2011 caretaker government
  

• 4 A Typology of Caretaker Periods

  To map the whole universe of caretaker cabinets in Belgium, we initially relied on the typology built by Baeselen and his colleagues (2014). To this we add two important insights. In this novel typology, we distinguish seven different types of caretaker periods (named A to G) depending on different scenarios: pre- and post-election periods, parliament being dissolved or in place, resignation accepted by the King, a new cabinet being sworn in. For the sake of clarity, we present this typology in a chronological sequence, hereby using the well-known example of the 2010-2011 political crisis (see Figure 1).
  The first type of a caretaker period (Type A in our typology) is initiated by the resignation of the cabinet, generally owing to internal dissent. In the Belgian political system, however, the resignation of the cabinet is effective only after the King/Queen has formally approved it. Usually, the prime minister presents his or her resignation to the sovereign who accepts this. Yet it may well happen that the King/Queen does not directly accept the resignation of the cabinet (which is increasingly the case in Belgian politics). Instead, (s)he may take a few days to consult political leaders or simply to ‘cool things down’. The resigning prime minister can then seize this opportunity to try and solve once more the rising political crisis that has led to the resignation of the cabinet. The reality of being at the brink of political deadlock may also be used as a leverage tool during negotiation talks. Later on, the sovereign may decide to finally accept the resignation (e.g. the Michel II cabinet in 2018) or to refuse it (e.g. the Leterme I cabinet in July 2008). During this reflection period, parliament remains in its full legal capacity. This type of caretaker period may seem anecdotic, but it actually concerns four different cabinet episodes during the observed time frame: it lasted for a total of 12 days between 2007 and 2020. Other countries, such as the Netherlands, have witnessed similar types of caretaker periods, where cabinets are labelled ‘demissionaire’ as soon as they have offered their resignation to the King/Queen (Otjes & Louwerse, 2014).
  The second type of a caretaker government concerns the interval that starts when – for whatever reason – the life of the cabinet is ended before the end of the legislature and whereby the King/Queen accepts this (Type B). This premature end of the cabinet may be due to a resignation of the government because of internal disagreements or after it lost a confidence vote in parliament. During this interlude, parliament remains in its full legal capacity. This kind of caretaker period typically ends with the dissolution of parliament (e.g. the Leterme II cabinet in 2010) or the swearing-in of a new government (e.g. the Leterme I cabinet in 2008). A telling example of this type of caretaker government lies in the period of 125 days of the Michel II cabinet that followed the acceptance of its resignation by the King on 21 December 2018 until the dissolution of the federal parliament on 25 April 2019. In total, this type of caretaker period lasted for 145 days between 2007 and 2020.
  The third type of caretaker period (type C) may chronologically follow the previous type and concerns the time segment between the dissolution of parliament and the conduct of new elections. Since the incumbent cabinet was already in caretaker mode at the time of the dissolution (e.g. the Michel II cabinet in 2018), the calling of new elections does not affect its actual working: it was already bereft of full control by parliament since its formal resignation. This delicate transition period is constitutionally bound and extends to a maximum of forty days. For example, in 2019 parliament was dissolved on 25 April and elections were conducted on 26 May. Overall, this type of caretaker period lasted for 68 days between 2007 and 2020.
  It may also happen that the coalition government does not resign before the end of its actual term (this happened twice during the 2007-2020 period) and enjoys its full powers at the time of the dissolution of parliament. This fourth kind (Type D) is different from the previous one in that the executive is not demissionary here but becomes caretaker just because the dissolved parliament can no longer control the government’s actions. This was, for instance, the case of the Di Rupo cabinet in 2014, which automatically became a caretaker government, and this from the day of the dissolution of the federal parliament onwards (24 April). The two episodes that correspond to this type of caretaker period together account for 70 days between 2007 and 2020.
  The fifth (Type E) period resembles the two previous types of caretaker periods but can be differentiated by the fact that it immediately follows upon new elections. The elections deliver a renewed set of representatives and may alter the balance of power in parliament. But the key issue for this period is that the new parliament has not been installed yet and therefore cannot exercise its controlling duties on the actions the cabinet undertakes in the meantime, no matter whether the cabinet was demissionary or not before the elections. During this post-election period, negotiations to form a new cabinet are initiated. In the recent political history of Belgium, no federal cabinet has been formed before the installation of the newly elected parliament. Usually, this period between the elections and the first gathering of the parliament lasts for three to four weeks (from 18 days in 2007 to 25 days in 2019). Within the time frame of our study, this period lasted for 91 days in total.
  The installation of a new legislature is a key moment in the lifetime of a caretaker government. Even if it does not directly impact the power of the executive, it means that parliament now re-enjoys its full legal capacity to control the actions of the caretaker cabinet. In addition, the new balance of power in the legislative branch can affect the number of seats on which the cabinet coalition relies. For instance, the minority cabinet Michel II could rely on fifty-two seats before the elections and merely thirty-eight afterwards. These electoral changes affect the capability of caretaker cabinets to pass new legislation. In brief, the sixth type of caretaker period (Type F) concerns the time span between the installation of a new parliament and the swearing-in of a new government. During this period, parties and their delegates discuss the new coalition formula, the content of the coalition agreement, the appointment of the new ministers and other key issues such as state and constitutional reforms. This extensive political agenda partly explains the length of these caretaker episodes, such as in 2010-2011, where it lasted for 518 days. This is by far the most frequent type of caretaker period in Belgian political history as it corresponds to no less than 1,079 days between 2007 and 2020.
  The last type of caretaker period (type G) corresponds to the interval between the swearing-in (the new prime minister and the new ministers take oath in the hands of the King/Queen) and the vote of confidence in parliament. Even if the new cabinet is installed thanks to a political agreement, it will only enjoy its full legal capacity after it has been formally approved by an absolute majority in parliament. This period may take a few days as the newly appointed prime minister needs to present his or her government declaration in parliament, followed by an investiture debate as well as a confidence vote. In 2014, it took no less than five days between the swearing-in of the Michel I cabinet (11 October) and the actual vote of confidence in the federal parliament (16 October). In total, this type of caretaker period accounts for 20 days during the time frame we investigate.
  In total, we observe no less than twenty-seven episodes of caretaker government during the period 2007-2020 in Belgium. Table 2 categorises these into seven types of cabinets. On average, a caretaker episode lasts for 55 days, ranging from small periods of merely two days (for instance, in 2020 between the Wilmès II being sworn in and the vote of confidence in parliament) to a period of 518 days in 2010-2011 between the installation of the newly elected parliament and the swearing-in of the Di Rupo cabinet. The most frequently employed caretaker period (72.66%) concerns episodes taking place after the renewal of parliament and before the appointment of a new cabinet. This comes as no surprise since the most important and delicate steps in the negotiations for the formation of a new coalition government take place during those moments.

  Table 2 Typology of caretaker periods

  Period type	Begin	End	Days (%)
  A	Resignation presented	Resignation accepted	12 (0.81)
  B	Resignation accepted	Parliament’s dissolution	145 (9.76)
  C	Parliament’s dissolution (Government is demissionary)	Elections	68 (4.58)
  D	Parliament’s dissolution (Government is not demissionary)	Elections	70 (4.71)
  E	Elections	Parliament’s installation	91 (6.13)
  F	Parliament’s installation	Swearing-in	1,079 (72.66)
  G	Swearing-in	Vote of confidence	20 (1.35)

• 5 Conclusion

  This research note was aimed at building upon the study of the world’s best known case of caretaker governments, with a view to improving the definition and the main characteristics of the concept. We developed a novel typology of the different caretaker periods based on the electoral calendar and on the moments when one of the two branches (executive or legislature) does not enjoy its full powers. This operationalisation allowed us to measure how Belgium was governed by a caretaker cabinet for no less than 1,485 days between 2007 and 2020. An exploration of the Belgian federal case for this time frame enables us to identify no less than seven different types of caretaker periods. Not only the total length of the caretaker governments but also each specific subperiod was precisely measured. In a next phase, this new typology and the distinction made between different types of caretaker periods can be highly useful in the evaluation of the quality of executive-legislature relations and of the overall quality and stability of the Belgian institutions and their democratic settings.
  Another possible next step is the validation of this typology in other countries and/or subnational contexts. In Belgium, for instance, it can be tested at the regional and community levels of governments. Many other parliamentary democracies, such as the Netherlands or Iceland, do also witness periods of caretaker government and could likewise make use of the distinction between different types of caretaker periods. The typology is particularly applicable in countries such as Australia and New Zealand, where the political leadership acts differently depending on the kind of transition period (for instance, before or after elections are announced – see, e.g., Simms, 2011). Furthermore, it could be analysed how governmental policy making and performance are affected during caretaker periods. To what degree are caretaker cabinets extending their own powers over time and across different types of caretaker periods? Last but not least, this custom of a caretaker government and its different appearances could be compared with the dissimilar political practice in countries such as Spain or Israel, where there is an automatic call for new elections after a given period of time has passed without forming a new cabinet. Similarly, the same custom could be compared with the situation in many Central European countries, where caretaker cabinets rather take the form of transitory or interim cabinets.

References

• Amorim Neto, O., & Strøm, K. (2006). Breaking the parliamentary chain of delegation: Presidents and non-partisan cabinet members in European democracies. British Journal of Political Science, 36(4), 619-643.

• Baeselen, X., Toussaint, S., Pilet, J.-B., & Brack, N. (2014). Quelle activité parlementaire en période d’affaires coura‍n­tes? Cahiers de l’ULB et du PFWB, 1.

• Beckman, L. (2007). The professionalisation of politics reconsidered. A study of the Swedish cabinet 1917-2004. Parliamentary Affairs, 60(1), 66-83.

• Boston, J., Levine, S., McLeay, E., Roberts, N. S., & Schmidt, H. (1998). Caretaker government and the evolution of caretaker conventions in New Zealand. Victoria University of Wellington Law Review, 28, 629-648.

• Bouckaert, G., & Brans, M. (2012). Governing without government: Lessons from Belgium’s caretaker government. Governance: An International Journal of Policy, Administration, and Institutions, 25(2), 173-176.

• Brans, M. (2012). Continuity and change in Belgium’s caretaker administration. European Political Science, 11, 102-107.

• Brans, M., Pattyn, P., & Bouckaert, G. (2016). Taking care of policy in times of crisis: Comparative lessons from Belgium’s longest caretaker government. Journal of Comparative Policy Analysis, 18(5), 1-16.

• Courtenay Ryals, C., & Golder, S. N. (2010). Measuring government duration and stability in Central Eastern European democracies. European Journal of Political Research, 49(1), 119-150.

• Davis, G., Ling, A., Scales, B., & Wilkins, R. (2001). Rethinking caretaker conventions for Australian governments. Australian Journal of Public Administration, 60(3), 11-26.

• Herman, V., & Pope, J. (1973). Minority governments in Western democracies. British Journal of Political Science, 3(2), 191-212.

• Hloušek, V., & Kopeček, L. (2014). Caretaker governments in Czech politics: What to do about a government crisis. Europe-Asia Studies, 66(8), 1323-1349.

• Hooghe, M. (2012). Does multi-level governance reduce the need for national government? European Political Science, 11, 90-95.

• Larsson, T. (1994). Cabinet ministers and parliamentary government in Sweden. In M. Laver, K. A. Shepsle (Eds.), Cabinet ministers and parliamentary government (pp. 69-182). New York: Cambridge University.

• Magone, J. M. (2000). Portugal. The rationale of democratic regime building. In W. C. Müller & K. Strøm (Eds.), Coalition governments in Western Europe. Oxford & New York, NY: Oxford University Press.

• McDonnell, D., & Valbruzzi, M. (2014). Defining and classifying technocrat-led and technocratic governments. European Journal of Political Research, 53(4), 654-671.

• Muller-Rommel, F., Fettelschoss, K., & Harfst, P. (2004). Party government in Central Eastern European democracies: A data collection (1990-2003). European Journal of Political Research, 43(6), 869-894.

• Otjes, S., & Louwerse, T. (2014). A special majority cabinet? Supported minority governance and parliamentary behavior in the Netherlands. World Political Science Review, 10(2), 343-363.

• Schleiter, P., & Belu, V. (2015). The challenge of periods of caretaker government in the UK. Parliamentary Affairs, 68, 229-247.

• Schleiter, P., & Morgan-Jones, E. (2009). Party government in Europe? Parliamentary and semi-presidential democracies compared. European Journal of Political Research, 48(5), 665-693.

• Simms, M. (2011). Westminster norms and caretaker conventions: Australian and New Zealand transition debates. In P. ‘t Hart & J. Uhr (Eds.), How power changes hands (pp. 94-107). Houndsmills: Palgrave Macmillan.

• Tiernan, A., & Menzies, J. (2007). Caretaker conventions in Australasia: Minding the shop for government. Canberra: ANU Press.

• Van Aelst, P., & Louwerse, T. (2014). Parliament without government: The Belgian parliament and the government formation processes of 2007-2011. West European Politics, 37(3), 475-496.

• Zafarullah, H., & Yeahia Akhter, M. (2000). Non-political caretaker administrations and democratic elections in Ban‍g­ladesh: An assessment. Government and Opposition, 35(3), 345-369.

Appendix

 

Table A.1 Detailed periods of caretaker governments (2007-2020)

Cabinet	Type	Period	Days
Verhofstadt II	D	02 May 2007 to 10 June 2007	39
Verhofstadt II	E	10 June 2007 to 28 June 2007	18
Verhofstadt II	F	28 June 2007 to 21 December 2007	176
Verhofstadt III	G	21 December 2007 to 23 December 2007	2
Leterme I	G	20 March 2008 to 22 March 2008	2
Leterme I	A	14 July 2008 to 17 July 2008	3
Leterme I	A	19 December 2008 to 21 December 2008	2
Leterme I	B	21 December 2008 to 30 December 2008	9
Van Rompuy	G	30 December 2008 to 02 January 2009	3
Leterme II	G	25 November 2009 to 27 November 2009	2
Leterme II	A	22 April 2010 to 26 April 2010	4
Leterme II	B	26 April 2010 to 07 May 2010	11
Leterme II	C	07 May 2010 to 13 June 2010	37
Leterme II	E	13 June 2010 to 06 July 2010	23
Leterme II	F	06 July 2010 to 06 December 2011	518
Di Rupo	G	06 December 2011 to 10 December 2011	4
Di Rupo	D	24 April 2014 to 25 May 2014	31
Di Rupo	E	25 May 2014 to 19 June 2014	25
Di Rupo	F	19 June 2014 to 11 October 2014	114
Michel I	G	11 October 2014 to 16 October 2014	5
Michel II	A	18 December 2018 to 21 December 2018	3
Michel II	B	21 December 2018 to 25 April 2019	125
Michel II	C	25 April 2019 to 26 May 19	31
Michel II	E	26 May 2019 to 20 June 19	25
Michel II	F	20 June 2019 to 27 October 2019	129
Wilmès I	F	27 October 2019 to 17 March 2020	142
Wilmès II	G	17 March 2020 to 19 March 2020	2
Total			1,485

Noten

• 1 The concept of ‘caretaker government’ is the most widely accepted term in the literature but may take different names in different countries; for example ‘interim government’ in Iraq, ‘demissionary cabinet’ in the Netherlands, ‘government in functions’ in Spain or ‘government of current affairs’ in Belgium.

• 2 Courtenay Ryals and Golder (2010) label these cabinets ‘new caretaker’ governments, while the type of cabinets studied in this note corresponds to what they call ‘continuation caretaker’ governments.

• 3 These authors are less clear about the presence of a caretaker cabinet in two other possible scenarios: (a) the period between the dissolution of parliament and the following general election (provided that the government still enjoys the confidence of parliament) and (b) the period between the announcement of an election and the dissolution of parliament (see, for instance, Simms, 2011 for a discussion of variants of political leadership during this period in Australia and New Zealand).

• 4 This period was labelled in the international media as the ‘world’s longest government formation period in modern history’.

• 5 For the clarity of the argument, we included the whole period of the Verhofstadt II cabinet (2003-2007) in this table. In the remainder of the research note, we focus solely on the 2007-2020 period.