Coalition cabinets are the norm in European politics, and this reality is unlikely to change given the increased fragmentation of party systems and decline of mainstream parties. The successful functioning of coalition cabinets is dependent on the extent to which coalition partners manage to stick to the agreed common coalition positions, instead of reverting to their own preferences. Inevitably, multiparty cabinets need collectively to delegate policymaking power to ministers from individual parties. As with any delegation act, this creates the risk that the agent (the minister) will follow the agenda of her party or her own personal preferences and not the policies of the coalition.
The academic literature has documented several oversight mechanisms that can be used by coalition partners to keep tabs on each other's ministers and avoid policy drift. These can be found at executive level (inner cabinets, cabinet committee(s), junior ministers), in the parliamentary arena (legislative review in committees, shadow committee chairs and coordination between majority parliamentary leaders) or in the extra-parliamentary arena (coalition agreements, coalition committees and party summits) (Strøm et al. Reference Strøm, Müller and Smith2010: 522).
A shadow committee chair describes a situation in which the parliamentary committee dealing with the policy jurisdiction of a certain department is headed by an MP from a different coalition party than that of the minister and that MP uses her position to monitor whether the legislation initiated by the minister or other actions of the minister follow the agreed coalition preferences. When ministerial drift is discovered, the chair can attempt to correct it or raise the issue with the coalition leaders. This article draws on an original data set, covering 11 cabinets and a 39-year period (1980–2018), to investigate the conditions under which committee chair allocation is used in Belgium by coalition parties to keep tabs on ministers of their cabinet partners. We analyse the extent to which the appointment of shadow committee chairs is driven by ideological distance and salience of the portfolio and whether there is a substitution effect between this practice and the usage of watchdog junior ministers and coalition agreements.
Our article builds on an expanding literature that analyses the allocation of committee chairs between parties and provides evidence for the theory that chairs are used strategically for shadowing purposes (Carroll and Cox Reference Carroll and Cox2012; Clark and Jurgelevičiūtė Reference Clark and Jurgelevičiūtė2008; Kim and Loewenberg Reference Kim and Loewenberg2005; Krauss et al. Reference Krauss, Praprotnik and Thürk2021). One alternative explanation is that the chair allocation process is used as compensation mechanism for coalition parties receiving fewer ministerial portfolios than they would be entitled to proportionally (Pukelis Reference Pukelis2016). However, the explanatory potential of this theory is bound to be limited given the fact that most cabinets do not deviate substantially from Gamson's Law in the distribution of portfolios (Falcó-Gimeno and Indridason Reference Falcó-Gimeno and Indridason2013).
Another model of chair allocation implies that the party which controls the relevant ministry would also receive the committee chair position to ensure it has full policy responsibility in that area. Martin Hansen (Reference Hansen2018) finds support for this model in the case of Denmark and links it to the main structural feature of Danish cabinets: their minority status, and their need to have some control over the committees in the absence of a supporting majority.
The introduction is followed by a discussion of the theoretical arguments and empirical evidence which inspired this study, and a presentation of the main hypotheses that will be tested. We then introduce the rationales for our case selection and the official rules that govern committee chair allocation and the chairs’ formal power in the Belgian legislature. Next, we present the details of the research design: the data sources, variables’ operationalization and methods. The fifth section discusses the results of the multivariate analyses followed by several robustness checks. The conclusions discuss the implications of the main findings and point to further directions of research.
Committee chair shadowing: previous evidence and new hypotheses
In most parliaments, committee chairs have influence over the committee agenda, are part of the legislature's coordinating bodies, receive substantial media attention (Chiru Reference Chiru2020; Chiru and Gherghina Reference Chiru and Gherghina2019; Fernandes et al. Reference Fernandes, Riera and Cantú2019; Gattermann and Vasilopoulou Reference Gattermann and Vasilopoulou2015; Rosenthal Reference Rosenthal2018) and are thought to play a key role in attempts to correct ministerial drift (Carroll and Cox Reference Carroll and Cox2012; Fortunato et al. Reference Fortunato, Martin and Vanberg2019). Nevertheless, some recent work has cast doubt over the shadowing argument regarding committee chairs. For instance, some authors argue there is not more scrutiny of cabinet legislation when the committee is chaired by a coalition partner MP – that is, a shadow chair – compared with when the chair is an MP representing the minister's party (Fortunato Reference Fortunato2019; Fortunato et al. Reference Fortunato, Martin and Vanberg2019). Instead, higher levels of scrutiny are registered when the committee is chaired by an opposition MP. However, evidence that coalition parties do use committee chairs to control their partners and shape legislation emerged from an analysis of German state parliaments (Krauss et al. Reference Krauss, Praprotnik and Thürk2021). Those findings suggest that it is more likely that a legislative proposal is changed at the committee stage if the committee chair belongs to the coalition partner.
In most Western European parliaments committee chairs have rather limited formal prerogatives, so their capacity to influence committee work in general is probably not very high (Sieberer and Höhmann Reference Sieberer and Höhmann2017).Footnote 1 Ulrich Sieberer and Daniel Höhmann (Reference Sieberer and Höhmann2017) found that there does not seem to be a correlation between committee chairs’ formal powers and the incidence of shadow committee chairs. Nevertheless, the existence of formal power for chairs within committee business is not a necessary condition for monitoring the activity of the minister. The broader literature cited above also does not necessarily rely on such assumptions about formal powers, but rather considers the committee chair simply to be the most elevated post within the parliament dedicated to the jurisdiction's policy. In the following paragraphs we present the arguments behind our four hypotheses, which focus on the ways in which ideological distance, portfolio salience and alternative mechanisms for monitoring the behaviour of ministers and implementing the coalition's agreed policies shape the likelihood of committee chair shadowing.
First, the risk of ministerial drift is substantially higher when the coalition partners have very different ideological positions and policy preferences. It is therefore reasonable to expect considerably larger monitoring efforts in such situations, and previous studies have confirmed this to be the case with respect to not only watchdog junior ministers (Lipsmeyer and Pierce Reference Lipsmeyer and Pierce2011), but also committee chairs. Dong-Hun Kim and Gerhard Loewenberg (Reference Kim and Loewenberg2005) in a longitudinal study of German cabinets, and Royce Carroll and Gary Cox (Reference Carroll and Cox2012) in a comparative study of 19 democracies have shown that the ideological distance between a minister's party and the other coalition parties increases the likelihood of shadowing by committee chairs. Instead, a study of intra-coalition politics in the Baltic states found that ideological distance affects the likelihood that committee chairs shadow ministers only indirectly. Thus, cabinet parties which are ideologically more distant from the other parties tend to receive fewer committee chair positions overall, which constrains their oversight capacities (Pukelis Reference Pukelis2018: 58–61). Because the assessment of ideological distance by coalition parties could happen either at party or at portfolio level, we formulate two hypotheses:
Hypothesis 1a: The larger the ideological distance between the minister's party and the cabinet average position on the left–right scale, the more likely she is to be shadowed by the committee chair.
Hypothesis 1b: The larger the ideological distance between the minister's party and the cabinet average position on the policy areas in the jurisdiction of the department, the more likely she is to be shadowed by the committee chair.
Second, policy salience influences not only which departments parties pursue as coalition payoffs (Bäck et al. Reference Bäck, Debus and Dumont2011; Ecker et al. Reference Ecker, Meyer and Müller2015; Raabe and Linhart Reference Raabe and Linhart2015) but also which committee chair positions they seek to obtain (Evans Reference Evans2020). Previous research has also shown that there is a higher likelihood of shadowing by junior ministers when the department deals with a more salient policy area (Lipsmeyer and Pierce Reference Lipsmeyer and Pierce2011). Although not tested directly previously, it would be reasonable to expect committee chairs to shadow departments that are more salient for the coalition as a whole, as agency drift in such cases involves substantive political costs because it hurts important policy interests of the parties or has high electoral costs attached. Conversely, coalition partners have few incentives to monitor each other if the policy area in question is politically irrelevant for them.
Hypothesis 2: The higher the saliency assigned by the cabinet parties to the policy areas in the jurisdiction of the department, the more likely it is that the minister will be shadowed by the committee chair.
Third, previous scholarship on intra-coalition politics has illustrated convincingly the frequent use of watchdog junior ministers as a key policing mechanism of coalition bargains (De Winter et al. Reference De Winter, Timmermans, Dumont, Müller and Strøm2003; Greene and Jensen Reference Greene and Jensen2016; Lipsmeyer and Pierce Reference Lipsmeyer and Pierce2011; Martin and Vanberg Reference Martin and Vanberg2011; Thies Reference Thies2001). However, the literature is much more ambivalent with respect to the relationship between appointing junior ministers and committee chairs for shadowing purposes. Thus, while some evidence would indicate a substitution effect, other studies found that the two practices are used simultaneously. Drawing on the German case, two studies (Kim and Loewenberg Reference Kim and Loewenberg2005; Thies Reference Thies2001) have hypothesized that watchdog junior ministers and committee chairs are used alternatively as mechanisms for keeping tabs on the ministers of coalition partners. Analysing data from the German Bundestag and federal cabinets covering a 32-year period (1966–98), Kim and Loewenberg (Reference Kim and Loewenberg2005) concluded that the higher the share of ministers shadowed by their junior ministers, the less frequent the appointment of shadow committee chairs. Somewhat related, Lanny Martin and Georg Vanberg (Reference Martin and Vanberg2011) found a substitution effect between using junior ministers for information gathering and the level of scrutiny by parliamentary committees in systems with strong parliamentary institutions.
However, other studies have revealed a positive association between the two practices: in young democracies, such as Lithuania and Latvia, when ministers are shadowed, they are shadowed by both junior ministers and committee chairs (Clark and Jurgelevičiūtė, Reference Clark and Jurgelevičiūtė2008: 637; Pukelis Reference Pukelis2018: 74). While using both types of shadowing mechanisms might be costly in terms of negotiating such positions and the efforts required for monitoring, their deployment in tandem is likely to be more efficient.Footnote 2 As speculated by Carroll and Cox (Reference Carroll and Cox2012: 223): ‘junior ministers may leak information to committee chairs, who can then slow down bills that have already reached the assembly, giving coalition partners time to negotiate their differences’. Such efficiency gains might therefore outweigh the opportunity costs induced by the simultaneous usage of multiple oversight mechanisms.Footnote 3
Hypothesis 3: Ministers facing hostile junior ministers will also be shadowed by committee chairs.
Fourth, and in a similar manner, we expect cabinets that have detailed coalition agreements to be keener to appoint shadowing committee chairs to make sure that these are implemented (Kim and Loewenberg Reference Kim and Loewenberg2005: 1108). The need for a comprehensive agreement may be due to ideological distance or reduced mutual trust, and as such it should call for as many oversight mechanisms as possible. Research on coalition agreements has shown that on average, in Western Europe, 90% of their content is policy related, the rest being taken up by procedural rules and office distribution (Strøm et al. Reference Strøm, Müller and Smith2010: 530). Having thorough policy deals agreed explicitly by the coalition in a contract would make monitoring easier (Bowler et al. Reference Bowler, Indridason, Bräuninger and Debus2016: 1275), implying that committee chairs would be better equipped to detect ministerial drift and to understand which amendments arising at the committee stage would go against the coalition's preferred position. In line with this complementarity argument, previous research has found that coalition agreements tend to be more detailed when junior ministers are appointed to keep tabs on ministers from different parties (Indridason and Kristinsson Reference Indridason and Kristinsson2013).
One counterargument to the multiple monitoring arrangement could be that governing parties which have put together detailed coalition agreements have already resolved their differences and therefore do not need to waste resources monitoring the behaviour of their partners. Nevertheless, such a scenario would imply a high level of mutual trust, which is a rather scarce resource, especially for parties that have not governed together before.
Hypothesis 4: The more detailed the coalition agreement the more likely it is that ministers will be shadowed by committee chairs.
Beyond the four hypotheses we control for the cabinet type, the minister's party seat share, and the number of seats held by the chair's party. The survival of minimum winning coalitions is much more dependent on the level of policy and office gains obtained by each of the coalition parties through their government participation than a surplus majority cabinet which, by definition, can even afford to see one of the partners depart (Chiru Reference Chiru2015; Damgaard Reference Damgaard, Strøm, Müller and Bergman2008). For this reason, we would expect minimum winning coalitions to be much more careful in monitoring the coalition bargain and therefore to use shadow chairs more frequently than surplus majority cabinets. Conversely, oversized coalitions usually include more parties than minimum winning coalitions, which increases the opportunities for shadowing.
Cabinet parties which control a large number of seats would have a smaller probability of being shadowed, given the proportionality of committee chair allocation. On the contrary, large parties would have a higher chance of shadowing smaller cabinet parties when they are the one appointing the committee chair – an intuition which should be captured by the number of seats held by the chair's party. Large parties have an advantage in Belgium not only because of the proportionality of the chair allocation process, but also because the size affects the order in which committee chairs are picked, as described below.
Case selection and the rules regarding committee chair selection and powers
Case selection
Belgium is an interesting but complex case for coalition research. Nearly all the 43 post-war governments have been coalition cabinets. In the period under analysis (1980–2018), due to soaring party fragmentation (effective number of parties between 6.8 and 9.1), cabinets contained four to six parties. Two cabinets were unconnected (Verhofstadt I and II, 1999–2007). About half of the coalitions were minimal winning, the other were surplus coalitions. Some had narrow majorities, other could count on large majorities. Coalition formation duration varies from 15 days to a world record of 541 days. The left–right polarization indexFootnote 4 varies from only 1.5 for the compact coalition of Martens V to 5.2 for the heterogeneous cabinet of Verhofstadt I. Finally, given the complexity of cabinet formation and maintenance, a comparatively large number of mechanisms have been installed to contain or solve interparty cabinet conflicts (De Winter et al. Reference De Winter, Timmermans, Dumont, Müller and Strøm2003).
Hence, this longitudinal analysis of nearly 40 years of coalition politics can help us to unveil factors that can explain the variation in the use of shadow committee chairs in Belgian coalition governance.
Formal rules regarding the allocation of committee chairs and their powers
In the Chamber of Representatives, the nomination of committee chairs follows several formal and informal rules. Chairs are allotted to parliamentary party groups in a proportional way (Règlement de la Chambre des représentants 2019: Art. 158), following the D'Hondt method.Footnote 5 In order to be recognized and subsidized as a group, a party needs at least five MPs (Règlement de la Chambre des représentants 2019: Art. 11.2). Both rules lead to an overrepresentation of large parties among committee chairs. After each election, the Conference of the Group Leaders defines the number of parliamentary positions allotted to each group. The parties choose the committees they want to chair in the order of their size: thus, the largest party has the first choice, followed by the second party, and so on. Ideally one would incorporate in the empirical modelling strategy the fact that the choice of committees happens in such a sequence, but unfortunately we lack this detailed information.
In one of the first plenary meetings (the third in 2019), the Speaker announces, upon proposal of the Conference of the Group Leaders, the MPs who will chair the permanent committees, while each committee nominates a first and second vice chair during its first meeting.Footnote 6
According to a comparative analysis of information acquisition, rewrite authority and timetable control in 31 democracies, Belgian parliamentary committees stand among the strongest in Western Europe (André et al. Reference André, Depauw and Martin2016: 111), which bodes well for their usage in the legislative review of ministerial proposals. Nevertheless, and as others have observed, the formal powers attributed exclusively to committee chairs in Belgium are at best moderate (Sieberer and Höhmann Reference Sieberer and Höhmann2017). The chair shares the committee agenda with the majority of the committee members and the Speaker (Règlement de la Chambre des représentants 2019: Art. 24). On the committee agenda, government bills (including budget bills) have priority over private member bills, but a simple majority of committee members could overrule a chair's agenda. Thus, agenda-setting is rather consensual between majority and minority committee members, and the committee as a whole vis-à-vis the minister. The chair decides on the scheduling of a meeting of the committee, delaying or suspending it. The chairs’ influence is also derived from their participation in the legislature's coordinating bodies, as chairs are often members of the House Bureau and of the Conference of Presidents which decides on the work schedule of the House (Règlement de la Chambre des représentants 2019: Art. 14; Art. 19).
Figure 1 shows that there is substantial variation in the allocation of committee chairs both between cabinets and across committees. Thus, the proportion of committee chairs shadowing ministers varies between 18.2% (Leterme I) and 60% (Dehaene I and II). The different number of committees is due either to the changes in the structure of the committees (e.g. the disappearance of the Business Committee) or because certain committees did not have a corresponding department in certain cabinets (e.g. the Committee for Institutional Reform).
The Allocation of Committee Chairs by Government and Committee
Data, operationalization of variables and methods
We collected the data regarding committee chairs’ party affiliation ourselves for the 1980s, early 1990s and for the most recent cabinet included in the sample. We draw on Steven Van Hauwaert and Chloé Janssen (Reference Van Hauwaert and Janssen2017) for the 1995–2014 data.
The first step of the analysis was to match the permanent committees and the ministries based on their policy jurisdiction. The full matching of the committees and portfolios for the 11 governments is displayed in Table A1 in the Online Appendix. The dependent variable is coded 1 if the committee that shares the same or a very similar policy jurisdiction as the department is chaired by a politician from a different coalition partner than the minister and 0 if the committee is chaired by an MP from the same party or from the opposition.
The ideological distance – party level variable indicates the absolute difference between the minister's party's left–right position and the seat-weighted average position of the cabinet. This was computed using the parties’ left–right scores retrieved from the Chapel Hill Expert Surveys (Bakker et al. Reference Bakker, de Vries and Edwards2015; Hooghe et al. Reference Hooghe, Bakker and Brigevich2010; Polk et al. Reference Polk, Rovny and Bakker2017; Ray Reference Ray1999; Steenbergen and Marks Reference Steenbergen and Marks2007).
The ideological distance – portfolio level variable indicates the natural logarithm of the absolute difference between the minister's party's positions on the policy dimensions that pertain to the jurisdiction of the department and the seat-weighted average position of the cabinet on the same dimensions, using the Manifesto Project (MARPOR) (Volkens et al. Reference Volkens, Burst and Krause2020) manifesto data.
Portfolio saliency indicates the seat-weighted sum of the cabinet parties’ assigned salience to the policy areas included in the jurisdiction of the department, using again MARPOR data. Our matching of MARPOR categories to policy areas follows past similar analyses (Bäck et al. Reference Bäck, Debus and Dumont2011; Hohendorf et al. Reference Hohendorf, Saalfeld and Sieberer2021) and is presented in detail in Table A2 in the Online Appendix.
Watchdog junior minister is a dichotomous variable coded 1 if a junior minister assigned to the department under consideration in that cabinet belongs to another coalition party from the minister. The data regarding this variable and the length of the coalition agreement were retrieved from Lieven De Winter and Patrick Dumont (Reference De Winter, Dumont, Bergman, Müller and Strøm2021).
The information regarding the cabinet type (minimal winning coalition v. surplus majority coalition), the number of seats held by the committee chair party and minister party were retrieved from the Parliaments and Government database (PARLGOV) (Döring and Manow Reference Döring and Manow2019). Table 1 presents the descriptive statistics for the variables in our sample. While 45% of the committee chair–minister combinations satisfy the shadowing condition, only 11% of the departments feature a junior minister.Footnote 7 As the dependent variable is dichotomous, we use a binary logistic regression for estimation. Moreover, because the decision of a coalition party to monitor the ministers of its partners might trigger similar behaviour from the latter actorsFootnote 8 we use robust standard errors clustered by cabinet.
Descriptive Statistics
What drives the use of committee chairs for shadowing in Belgium?
Table 2 reports the results of two logistic regressions. While the first regression draws on the full sample, the second one includes only the committees chaired by MPs who are members of cabinet parties.Footnote 9 Virtually all independent variables and controls behave similarly in both models. Irrespective of model specification, Hypothesis 1a is corroborated: the larger the left–right distance between the party of the minister and the mean cabinet position, the greater the likelihood of having a committee chair from a different coalition partner. The magnitude of this effect is illustrated in Figure 2.
Ideological Distance (Party Level) and Likelihood of Minister Being Shadowed
Determinants of Chair Shadowing Ministers (logistic regression)
Notes: Significance at * p < 0.10, ** p < 0.05, ***p < 0.01. Cell entries are odds ratios. Robust standard errors clustered by cabinet in parentheses.
Against our expectations, ideological distance at the level of the portfolio and the salience of the portfolio do not seem to influence the allocation of committee chairs in either of the models. Future studies could attempt to test whether this non-finding is due to relying on manifesto data designed first and foremost to capture the saliency of policy issues for parties and not their positions (Laver Reference Laver and Laver2001).
Hypothesis 3 is corroborated as there is no substitution effect for watchdog junior ministers. On the contrary, ministers that face ‘hostile’ juniors are very often also confronted with shadow committee chairs. Ministers serving in minimum winning coalition cabinets have a higher likelihood of being shadowed by committee chairs than ministers who are part of surplus majority cabinets. The magnitudes of these effects are illustrated in Figure 3.
Watchdog Junior Ministers, Cabinet Type and Likelihood of Minister Being Shadowed
We find no evidence for Hypothesis 4. Thus, ministers from cabinets that were negotiated for a long time and are based on more detailed coalition agreements have the same chance of being shadowed as their counterparts from cabinets with shorter coalition agreements.
A possible explanation concerns the fact that in Belgium coalition agreements are taken very seriously, irrespective of their precision (Moury Reference Moury2009; Timmermans and Moury Reference Timmermans and Moury2006). A minister can only send a bill to the House and the relevant committee when it has the approval of the full council of ministers. Hence, verification of ministers’ bills’ conformity to the coalition agreement is exercised at executive level, and any controversial matter will already be streamlined by the inner cabinet (‘Kerncabinet’, including the PM and all vice-PMs) (De Winter and Dumont Reference De Winter and Dumont2006).
Based on Model 1 in Table 2, Figure 2 plots the predicted probability of committee chair shadowing depending on the level of ideological distance between the minister's party and the cabinet mean left–right position, weighted by the seat contribution of each party to the coalition. The figure indicates that when the position of the minister's party is virtually identical to that of the cabinet, the likelihood of shadowing is 24%, whereas at the maximum distance observed in the sample (2.5), the probability that the committee chair is allocated to another coalition partner is 70%, all other things being equal.
Drawing on the same Model 1 in Table 2, Figure 3 plots predicted probabilities for two other independent variables. The left-side panel of Figure 3 illustrates that, indeed, watchdog junior ministers have a positive impact on the likelihood of shadowing, and the effect is relatively large. Thus, departments which include junior ministers from coalition partners are also much more likely to be shadowed by committee chairs than ministers who work with juniors from their own party, or who lack deputies. In the former case the probability of shadowing is almost 21% higher than in the latter case.
The right-side panel of Figure 3 indicates that ministers of surplus majority cabinets have a 37% chance of being shadowed by committee chairs, whereas for minimum winning coalition ministers the probability is 52%.
Robustness checks
The first robustness check was to rerun the model with a multinomial logistic regression having as DV a variable with three categories: chair allocated to opposition party, chair allocated to the party of the minister and shadowing chair. The results of these regressions (having as reference category the opposition committee chairs) are reported in Table A3 in the Online Appendix. Hypothesis 1a is again corroborated: a greater ideological distance between the minister's party and the cabinet leads to a higher likelihood of appointing a shadow chair than one from the ranks of the opposition parties. Watchdog junior ministers also seem to have a positive effect on appointing a shadow chair, but the variable does not reach conventional levels of statistical significance.
The length of cabinet formation can be considered another proxy for the level of disagreement between the coalition parties and the need to police the coalition bargain. In the Belgian case, this formation duration is strongly correlated with the comprehensiveness of the coalition agreement, thus we ran an alternative model in which the latter variable was replaced by the former. As can be seen from Table A4, this variable did not make a difference and the main results remained virtually identical, both in terms of direction and magnitude of effects.
Another robustness check was to replace the average portfolio saliency measure with a variable indicating the portfolio saliency only for the coalition party that actually nominated the committee chair. The results are reported in Model 1 in Table A5 in the Online Appendix, and they show this alternative operationalization does not make a difference. Moreover, we also replaced the departmental salience variable with a continuous variable based on the assessment of the Belgian case in James Druckman and Paul Warwick's (Reference Druckman and Warwick2005) expert survey on portfolio salience. As shown by Models 2 and 3 in the same table, the main results are robust to this change.
Last but not least, for six of the legislatures (1988–2010) we analysed whether there is complementarity between the use of shadow committee chairs and monitoring via parliamentary questions. This was motivated by the findings of recent research which has shown that coalition parties in the UK (Martin and Whitaker Reference Martin and Whitaker2019) and Germany (Höhmann and Sieberer Reference Höhmann and Sieberer2020) also use parliamentary questions to keep tabs on their partners’ ministers. Drawing on data from the Belgian Comparative Agendas Project (Walgrave et al. Reference Walgrave, Joly, Sevenans, Baumgartner, Breunig and Grossman2019), we computed the shares of oral questions and interpellations addressed to each department by MPs belonging to coalition parties other than the party appointing the respective minister.Footnote 10 We found no association between this variable and the use of shadow committee chairs.
Conclusions
This article contributes to the expanding literature on oversight mechanisms used by cabinet coalition partners to avoid ministerial drift by re-testing in a more fine-grained manner previous findings about the roles of ideologically distance and portfolio saliency in the use of committee chairs for shadowing purposes, and by revealing the relation between this practice and other monitoring arrangements. Thus, our analysis of this unexplored field of Belgian coalition governance indicates that shadowing ministers by allocating committee chairs to other coalition parties is present but varies both between cabinets (from 18% to 60%) and across committees.
Our results reveal that ideological distance matters at party but not at portfolio level: the greater the left–right distance between the party of the minister and the mean cabinet position the higher the likelihood of having a committee chair from a different coalition partner.
On the contrary, ideological distance at the level of the portfolio does not seem to influence committee chair shadowing. This might be because for such a measure to work it would imply a very demanding level of detailed knowledge on each party's respective policy preferences from the coalition leaders. Alternatively, the non-finding might be due to the measurement of this variable: manifesto data designed mainly to capture the saliency of policy issues for parties and not their positions. Future studies relying on different data to measure ideological distance at the level of portfolio and portfolio salience could also consider the interaction effect between the two variables as, theoretically, incentives to monitor each other are at a maximum if the coalition parties are ideologically far apart in a salient policy area.
Surprisingly, the saliency of the department to the cabinet parties does not affect the probability of committee chair shadowing, irrespective of the operationalization: mean weighted saliency for the coalition parties, saliency for the party that actually nominated the committee chair, or as perceived by experts.
We find no substitution effect for watchdog junior ministers, as ministers that face ‘hostile’ juniors are very often also confronted with shadow committee chairs. Ministers serving in minimum winning coalition cabinets have a higher likelihood of being shadowed by committee chairs than ministers who are part of surplus majority cabinets. On the other hand, we could not corroborate the hypothesis that ministers from cabinets that are based on more detailed coalition agreements have a higher chance of being shadowed than their counterparts from cabinets with shorter coalition agreements. The results of these multivariate analyses were confirmed by a number of robustness checks.
Further research could explore the institutional and partisan conditions under which coalition partners decide to use in tandem watchdog junior ministers and shadow committee chairs to avoid ministerial drift. A reasonable expectation would be that this would not happen in systems with weak parliamentary committees (Martin and Vanberg Reference Martin and Vanberg2011). Another hypothesis worth exploring is whether the allocation of committee chairs is also used as a compensation mechanism for the coalition parties’ anticipated future electoral losses. A survey experiment of Irish legislators illustrated that political elites are willing to compensate a party that is likely to be penalized by voters for its cabinet participation by allotting it more cabinet seats (Martin Reference Martin2018).
Supplementary material
To view the supplementary material for this article, please visit https://doi.org/10.1017/gov.2021.27.
Acknowledgements
Mihail Chiru's work was supported by UCLouvain through the ‘MOVE-IN Louvain’ Incoming Postdoctoral Fellowship, co-funded by the Marie Curie Actions of the European Commission.
