The Recovery and Resilience Dialogues: Cheap Talk or Effective Oversight?

The recovery and resilience dialogues were introduced by the regulation of the Recovery and Resilience Facility, and the first of such dialogues took place in May 2021. The European Parliament invites the Commission, approximately every two months, to exchange views on matters relating to the national recovery and resilience plans and progress in their implemen‐ tation. Through an analysis of an original dataset composed of the questions asked by the MEPs in the 10 dialogues held between May 2021 and April 2023, this article provides a systematic empirical assessment of the European Parliament’s capacity to hold the Commission accountable. Drawing on the literature on the economic and monetary dialogues and adapting the operationalisation of key variables to the new instrument, this article shows that the recovery and resilience dialogues are an effective instrument for information exchange and debate, but they serve as a weak instrument of political accountability. Additionally, it casts new light on significant differences between MEPs: South and East European members are considerably more active than members from Northern Europe. At the same time, parliamentarians only occasionally ask questions targeting other member states.


Introduction
In July 2020, the European Council reached an agreement on NextGenerationEU (NGEU): the Commission was empowered to borrow up to €672.5 billion (in 2018 prices) from financial markets to enable the member states to recover from the pandemic crisis and to push their digital and green transitions.The key instrument of NGEU, the Recovery and Resilience Facility (RRF), allocated up to €312.5 billion in grants and €360 billion in loans to member states.Although the EU has been borrowing from the markets since the 1950s, NGEU combines traditional features with some novel ones (Hodson & Howarth, 2023).Significantly, it is financed directly by the EU rather than through transfers from member states, and the amount of money the European Commission has been authorised to borrow is exceptionally large.The RRF also has tangible redistributive effects, with those countries most hit by the pandemic receiving a larger quota of resources (i.e., Italy in absolute terms, Greece as a share of its GDP).
To receive the EU funds, member states submit national recovery and resilience plans (NRRPs) to the European Commission, which assesses them and proposes its recommendations to the Council for final approval.Payments are subsequently managed by the Commission and made in successive instalments, pending a satisfactory assessment of the member state's progress toward agreed milestones and targets.The RRF Regulation concluded in February 2021 tasked the European Parliament (EP) with scrutinising the RRF and its implementation.Through the recovery and resilience dialogues (RRDs), the European Commission is heard and questioned by MEPs.To what extent are the RRDs ensuring the accountability of the Commission before the EP? Drawing on a classic definition, we understand political accountability as the requirement for public officials to justify their conduct before an accountability forum (typically a parliament), be assessed for their (past) actions, and possibly face consequences (Bovens, 2007).Broadly speaking, the Commission is directly accountable to the EP: its president and the College of commissioners are voted into office by the EP, which also has the power to table a motion of no confidence, forcing its collective resignation.The RRDs are an additional accountability tool that allows the MEPs to ask the Commission to justify its conduct or change its actions.In turn, commissioners shall publicly defend their actions or promise MEPs that they will change their behaviour.Although commissioners cannot be individually dismissed, and the possibility of collective removal is highly unlikely, the Commission's behaviour can be publicly exposed to negative media coverage and publicity and, therefore, sanctioned with reputational and political costs.
Previous studies have shown how parliamentary dialogues, particularly the economic dialogues and monetary dialogues, have been used by the EP to hold executive actors in economic governance and monetary policy accountable, also highlighting the limits of such instruments (e.g., Chang & Hodson, 2019;Kluger Dionigi, 2020;Maricut-Akbik, 2022).Building on this literature, we provide a comprehensive analysis of the ten RRDs held from May 2021 to April 2023.Our main goal is to empirically assess the strength and type of accountability, classifying all the questions asked by MEPs based on their content and the answers provided by commissioners based on their level of detail.In particular, we investigate three dimensions of accountability: the strength of the EP's accountability efforts, the level (national, supranational, or transnational) to which the EP's accountability is directed, and the change (or lack thereof) in the MEP's oversight over time.To our knowledge, our analysis provides the first empirical exploration of these dialogues.As such, rather than systematically testing hypotheses or causal mechanisms, it aims to generate new insights and provide original empirical data for further investigation by future research.
On the strength of accountability, we find that RRDs are useful instruments for MEPs to gather information and publicly debate issues related to the implementation of the RRF.However, they are weak instruments of accountability, as the Commission is only occasionally asked to justify or change its actions.
Regarding the level to which the accountability is directed, the RRDs are mainly used by the EP to inquire about the RRF rules and challenges for the EU as a whole, focusing less on the member-state level.To an extent, this is understandable, as national parliaments are also responsible for scrutinising their national governments, which are in charge of implementing the NRRPs.However, MEPs could-and, arguably, should-focus more on how EU funds are disbursed and spent across the member states, also scrutinising the implementation of the RRF in countries rather than their own.
Finally, looking at the longitudinal change in MEPs' oversight, we find evidence that, over time, the strength of their scrutiny does not change, their focus on nationallevel issues increases, and the commissioners become slightly more explicit in their answers.
We complement the assessment of the RRDs as accountability tools with a second line of inquiry, i.e., an analysis of which MEPs take the floor more frequently in the dialogues.Although we address this question only in an exploratory fashion, our findings reveal that MEPs from Southern and Eastern Europe are particularly active in the dialogues, while MEPs from Nordic Europe especially are underrepresented.This suggests that the redistributive aspects of the RRF mainly activate those MEPs from the countries which mostly benefit from the EU financial resources for post-pandemic recovery.This finding may cast some doubts on the legitimacy of the EP as the EU accountability forum.At the same time, it shall be recalled that the selection of MEPs is driven by various factors: most fundamentally, their expertise as (shadow) rapporteurs on legislation on post-pandemic recovery.Therefore, a national "bias" could have a solid functional justification (i.e., as the most engaged MEPs in the dialogues are, effectively, policy experts).
Our analysis of the RRDs not only casts light on the role of the EP in post-pandemic recovery, but it may also be illustrative of the difficulties encountered by parliaments to exercise effective oversight of executive actions in (post-)crisis situations.According to scholarship going back to Schmitt (Ginsburg & Versteeg, 2021, p. 1498), "Emergency governance is…executive governance."Parliaments play, at best, a reactive role and are ill-equipped to oversee executive actions (see also Bressanelli & Chelotti, 2019).In the post-pandemic EU system, the "NGEU show" is run by executive actors: at the EU level, the Commission and the Council (assessing implementation and authorising payments); at the member state level, the national governments (implementing the NRRPs).The RRDs are the main tool through which the EP has attempted to exercise its scrutiny over the actions of the supranational executive, and, as our empirical analysis demonstrates, it has only held it accountable to a limited degree.
This article is structured as follows.Section 2 introduces the RRDs and discusses them regarding other dialogues previously established by the EP.Section 3 outlines the conceptual framework and our expectations.Section 4 illustrates the research design and the coding procedure, while Section 5 presents and discusses the empirical analysis.Section 6 concludes.

The Recovery and Resilience Dialogues
Negotiating the RRF under the ordinary legislative procedure, the EP managed to expand its powers to scrutinise the implementation of the RRF (Fasone, 2022;Bressanelli, 2022).According to Regulation (EU) 2021/241 (2021), the Commission is expected to submit an annual evaluation report on the implementation of the RRF to both the EP and the Council.The NRRPs submitted by member states to the Commission and the proposals for Council implementing decisions must be sent to both the EP and the Council "simultaneously and on equal terms."The EP must also "simultaneously" receive information that the Commission relays to the Council, while its competent committee is informed about the "relevant outcomes of discussions" in Council preparatory bodies.The Commission shall also keep the EP updated on the fulfilment of the relevant milestones and targets during the implementation phase.Finally, the procedure to suspend payments to member countries requires the Commission to keep the EP fully informed in a timely manner, and allows the EP to invite the Commission to explain the motivations behind any proposal for suspension (or for lifting such a suspension).
An important institutional innovation introduced by the RRF Regulation has been the establishment of the RRDs.Consistent with the assumption that crises act as a trigger of institutional change (see Bressanelli & Natali, in press), the EP successfully managed to insert RRDs into the regulation as an instrument of accountability, one which was not originally included in the legislative proposal presented by the Commission.Thus, Art. 26 of the RRF Regulation allows the EP to invite the Commission, every two months, to exchange views on the implementation of the RRF, about the state of recovery, resilience, and adjustment capacity in the EU; the assessment of the NRRPs; the main findings of the evaluation report presented by the Commission; the status of fulfilment of the milestones and targets in the NRRPs; payment, suspension, and termination procedures; and any other relevant information and documentation provided by the Commission concerning the implementation of the RRF.The regulation stresses that the views expressed in the dialogues, together with the relevant resolutions voted by the EP, shall be duly considered by the Commission.
The RRDs take place before a joint meeting of two parliamentary committees-EP's Committee on Budgets (BUDG) and the Committee on Economic and Monetary Affairs (ECON).An ad hoc body (a working group composed of 27 full members and 14 substitutes, representing all political groups and mainly-but not only-recruited from the BUDG and the ECON Committees) is responsible for the preparation and the follow-up of the RRDs.
The instrument of the "dialogue" is not new, as it has already been used by the EP in other contexts.
The "economic dialogue" has been introduced by the legislation reforming economic governance, the sixand the two-pack (i.e., Bressanelli & Chelotti, 2016;Kluger Dionigi & Koop, 2019).Through it, the ECON Committee invites representatives of the member states, the European Commission, the president of the Council, and the president of the Eurogroup to discuss economic and policy issues.In the "monetary dialogue," instead, the president of the European Central Bank (ECB) appears on a quarterly basis before the ECON Committee.He/she delivers an introductory statement on the EU's economic and financial outlook and the ECB's decisions to MEPs, who then address their questions to him/her (see Moschella & Romelli, 2022).
Compared to the other types of "dialogues," some specificities of the RRDs are worth emphasising.First, unlike economic dialogues, where the president of the Council and representatives of member states can also be invited, only the Commission is expected to participate.While it is for the national governments to submit the NRRPs and ensure their implementation, the EP is only empowered to address questions to the Commission (Dias Pinheiro & Dias, 2022).Incidentally, the EP is not involved in the approval of the NRRPs or the adoption of the implementing decisions that authorise financial contributions.Second, the RRF regulation gives a broad remit to the RRDs, as it allows the EP to discuss "any relevant information…concerning the implementation of the RRF."Even if, in practice, the agenda of economic dialogues has also been quite broad (Maricut-Akbik, 2022), their regulations are, in theory, more stringent.

Conceptual Framework
To what extent, and how, does the EP hold the Commission accountable within the RRF framework?Parliamentary questions are one of the few mechanisms at the EP's disposal to control the Commission's exercise of power.They allow MEPs "to check, verify, inspect, criticise, or challenge [its] activities" (Maricut-Akbik, 2022, p. 2).In what ways are they used in the context of the RRDs to inspect, criticise, and challenge the decisions of the Commission?
In this article, we break down accountability into three analytical dimensions.First and foremost, we analyse the strength of the EP's oversight of the Commission.Previous work has shown that the accountability assured by the monetary and economic dialogues was, overall, limited.The monetary dialogue provided greater transparency and legitimacy to the ECB's decisions, but space for significant improvement existed-e.g., reducing the number of policy issues discussed or improving the cooperation among MEPs (Chang & Hodson, 2019;Maricut-Akbik, 2021).Kluger Dionigi (2020) measured the quality of the scrutiny in economic dialogue and showed that most of the MEPs' questions were unfocused and hardly directed to scrutinise the actions of EU institutions.By looking at the strength of the questions asked by MEPs and the responsiveness of EU executive actors, Maricut-Akbik (2022) reached similar conclusions: the EP has been able to hold EU institutions accountable only partially-with the scrutiny being stronger in the case of the Eurogroup than of the Commission or the Economic and Financial Affairs Council (ECOFIN).
We devote most of our efforts to operationalising and measuring the strength of these accountability efforts.Does the EP act through softer (e.g., requesting more information) or stronger (e.g., demanding specific changes of policy or behaviour) forms of oversight?At the same time, accountability also includes assessment of the quality of the Commission's response-i.e., the types of answers provided by the commissioners.Our assessment of the strength of the EP's oversight in the RRDs employs and adapts Maricut-Akbik's (2021) framework and methodological approach, which is illustrated in Section 4.
Second, we examine the level to which the MEPs' questions are addressed.On the one hand, MEPs represent EU citizens, not their national constituencies.One of the strongest arguments in favour of expanding the EP's powers is the legitimacy gap created every time EU integration has deepened.As more and more powers are assigned to the EU institutions, the national accountability chain is weakened, if not severed.The EP hasoften successfully-exercised normative pressure to fill this void (Rittberger, 2003).As such, we should expect that, in the case of RRDs, MEPs' oversight would be primarily directed to decisions and policies that concern the EU as a whole.
On the other hand, we also know that MEPs are very much interested in, and have great knowledge of, their domestic politics.There might be many reasons why they would ask questions about their own country's NRRP-not least because they can score points in their national polity by defending or criticising the situation at home.Evidence from the economic dialogues indeed reveals that MEPs are more likely to question ministers from their own member states (Brack & Costa, 2019;Chang & Hodson, 2019;Kluger Dionigi, 2020).There is also a third possibility: that parliamentarians pose questions that inspect what happens in other member states.For instance, MEPs attentive to the rule of law might look into the NRRPs of Hungary or Poland.The debtor/creditor distinction could apply at this level: countries from Northern Europe might have strong incentives to examine the design and implementation of NRRPs of highly indebted countries.We therefore aim to evaluate whether questions are directed at the EU, the national, or the transnational level.
Finally, we add a temporal dimension to our analysis.We cover 10 dialogues from May 2021 to April 2023.During this time, it is likely that the kind of accountability exercised by the EP changed.This could be because MEPs have become more experienced in the RRDs process, format, or content.They might have learned how to better review and challenge the commissioners, or they might have acquired greater knowledge of the dossiers.Alternatively, this could be because the dialogues simply happened at the different stages of the RRF cycle, e.g., MEPs initially asking questions related to the EU level (e.g., the rules of the games, the methodology used by the Commission in the assessment of the plans) and then inquiring about the implementation of specific projects in the member states once the NRRPs were underway.Not only the questions of MEPs but also the answers of the commissioners could/should change over time.Commissioners are likely to become more thorough (or "explicit," in the terminology we employ) in their answers as they acquire greater expertise with the RRF and the way RRDs are conducted.
Therefore, our research aims to assess the EP's accountability of the Commission in the RRDs by looking at the (a) strength, (b) level, and (c) change over time of these accountability dimensions.To this main line of enquiry, we add a second, exploratory one, where we investigate which MEPs take the floor in the RRDs.Given that the time available for these executive-legislative exchanges is scarce, only a handful of parliamentarians are able to ask questions.Exploring their profiles can shed greater light on the type of oversight provided by the EP.MEPs' engagement in these fora can be explained by a plurality of reasons.For instance, the EP political groups have an incentive to actively participate in the RRDs so that their voice can be heard by the commissioners, or they can signal their political activism to their constituencies (Bowler & Farrell, 1995).
In this article, we preliminarily assess two aspects of MEPs' individual profiles, i.e., their nationality and partisan affiliation.Most importantly, we want to determine if there is any systematic pattern in the nationality of the parliamentarians who take the floor.While the null expectation is that the share reflects the country number of seats in the joint BUDG-ECON Committees, an interesting question is whether there is a significant difference between Northern and Southern countries, or better, between countries that received the most (e.g., Italy, Spain, Greece, Romania) and the least (e.g., Netherlands, Denmark, Sweden) in terms of financial resources from NGEU.We expect that MEPs from the latter group-who are likely to be net contributors to the programme-will have greater incentives to participate in the dialogues and hold the Commission accountable for the running of the NRRPs.On the other hand, Kluger Dionigi (2020) has shown that opposite trends are at work in the economic dialogue: MEPs from debtor countries are more likely to ask questions than MEPs from creditor countries.
The allocation of speaking time in the RRDs is also likely to be related to the specialist role of MEPs in the joint BUDG-ECON Committees (i.e., Proksch & Slapin, 2011, p. 61).More senior MEPs (i.e., chairs or vice-chairs of the BUDG and ECON Committees) and policy experts (i.e., rapporteurs or shadow rapporteurs of the regulations on the RRF or REPowerEU) are likely to be in the best position to inspect and question the actions of the Commission.As such, the probability of them being selected by their political group to take the floor in the RRDs may well be higher.

Research Design
This section illustrates the methodological choices adopted to analyse the RRDs (for additional details, see the Supplementary File).We analyse the type and strength of the EP's accountability through a "political claims analysis" that involves the qualitative analysis of the claims made by specific actors in the context of interest (Koopmans & Statham, 1999, p. 6).The basic unit of analysis is the "claim."A claim can be defined as a "purposive unit of strategic or communicative action in the public sphere" (de Wilde et al., 2014, p. 7).In our approach, a claim is "a sentence or a set of sentences on a particular topic" (Maricut-Akbik, 2021, p. 551) and can be either a question by the legislative actors (in our case, the MEPs) or an answer by the addressees (in our case, the EU commissioners).
Critically, claims cannot be confused with "interventions" as formally defined during legislative oversight procedures.Usually, each intervention by a legislative or executive actor contains multiple claims, which can be identified both through explicit statements (i.e., when they enumerate their questions/answers) and/or changes in topics (Maricut-Akbik, 2021, pp.13-14; see Supplementary File).Therefore, the first step of political claims analysis is isolating the single questionsclaims and answers-claims.For example, this quotation comes from an intervention by an MEP in the RRD of 12 September 2022: On macroeconomic considerations, I'd like to pick up on what Commissioner Gentiloni said.You're right, we need to look at necessity, we need to coordinate budgetary and monetary policy.My question is the following: how can that be done at the EU level?Coordinate budgetary and economic or monetary policy?How can we strengthen these policies to fight inflation, to prevent our economies from falling into recession?Another question.The decision on a global minimum tax: what is the state of play on the debating Council?We know that at the most recent Ecofin Council, some Member States said that they would like to see enhanced cooperation here.That is my question: what is the current update on that debate?Thank you.
Through the specification of the MEP and the shift of topic, we can notice that there are two separate question-claims.The first is about budgetary and monetary policy coordination and the second is about the Council debate on the OECD global minimum tax.From now on, we use the words "questions" and "answers" to indicate "question-claims" and "answer-claims."After isolating questions and answers, a series of features are investigated for each.
To assess the strength of the EP's accountability through RRDs, we employ the Q&A approach to legislative oversight (hereafter, the Q&A approach), which is a framework provided by Maricut-Akbik (2021) to analyse parliamentary oversight procedures.Such an approach assumes that to assess the effectiveness of parliamentary questions, the questions asked by an accountability forum should be analysed together with the answers from the executive actors (Maricut-Akbik, 2021, p. 543).In particular, the Q&A approach identifies five types of questions that the legislative actors can ask the executive: (a) request information about a decision taken; (b) justify a decision or conduct; (c) change a decision or conduct; (d) adopt sanctions toward certain actors; and (e) request policy views, when the question does not contest anything about the past conduct of the addressees.In addition to the original framework, we have added a sixth category of (f) irrelevant questions, for questions on policy issues that fall outside the scope of the dialogues, as defined by Art. 26 (cf.Section 2).
Depending on how the addressee engages with the type of question, the executive's replies are classified as (a) explicit, (b) intermediate, and (c) non-reply.Explicit replies fully engage with the core of the question by presenting an exhaustive answer, intermediate answers are partial and/or incomplete, and non-replies do not engage with the question at all (Maricut-Akbik, 2021, pp.23-25; see Supplementary File).
To assess the level to which the EP's oversight is directed, we have introduced a new variable called "level," which captures whether a question focuses on the EU level, the national level (when MEPs ask a question about their own country), or the transnational level (when MEPs ask a question about a different country from their own).To scrutinise the temporal dimension, every claim is linked to the RRD in which it was made.
Finally, to make a preliminary assessment of the second line of inquiry (about who takes the floor during the RRDs) we compare the share of interventions with the share of seats in the joint committee both by political group and nationality.Information on the composition of the joint committee and the MEPs' political affiliation, nationality, and roles-as (vice)chair of either the BUDG or the ECON Committees or (shadow) rapporteur on legislation on the RRF or REPowerEU-were retrieved from the website of the EP (European Parliament, 2023; last accessed on 23 June 2023).
In practical terms, the coding procedure consisted in watching the video recordings of the RRDs available on the EP's website-European Parliament Multimedia Centre (europa.eu)-and,following our codebook adapted from Maricut-Akbik (2021), creating a dataset where every row includes a single question-claim by an MEP and the connected answer-claim by the commissioner(s).As additional support for the coder, every RRD was automatically transcribed with speech-to-text software.Whenever questions were asked in a language not known to the coders, the English translation provided by the EP was used.
In order to improve the validity of the coding, at the beginning of the process, each author independently coded one of the dialogues, following the Maricut-Akbik (2021) codebook.Later, we ran a "collective coding" session in which we compared our coding choices to clarify the interpretation of the various parameters and adopt modifications to the codebook.As we soon realised that the classification of some questions could vary a lot depending on the coder's interpretation of the tone and the context, we compiled and shared a vademecum (see Supplementary File) to interpret the codebook most coherently.
After this step, every author has coded one-third of the dialogues.They have been assigned in such a way that no one had more than two consecutive dialogues to give every coder a wide variety of periods and topics.After coding all the 10 dialogues, we conducted an inter-coder reliability test that consisted of (a) extracting a stratified sample of claims, (b) having such sample independently coded by two different coders, and (c) measuring the level of agreement with the following indices: percentage agreement, Krippendorff-alpha, and Gwet's AC 2 .Overall, for the three indexes, the scores reflect high reliability according to the current literature.The percentage agreements for question type, answer type and focus are 73.48%,77.27% and 92.42%, respectively.The Krippendorff-alpha scores are 0.7054, 0.5870, and 0.7820, respectively.The partly unsatisfying result for the Krippendorff-alpha on answer type (below the minimum score of 0.667 indicated by Krippendorff, 2004) despite the high percentage agreement might be related to the skewed distribution of the variable (Feinstein & Cicchetti, 1990;Gwet, 2015).For this reason, we also tested for an index that is less sensitive to such a feature, Gwet's AC 2 .With that index, the scores are 0.9430, 0.7810, and 0.9345, respectively, thus supporting the overall robustness of the intercoder reliability of the dataset (see Supplementary File for additional details).

Empirical Analysis
The analysis covers the 10 dialogues held between May 2021 and April 2023.Before delving deeper into the dialogues, it is useful to provide some information about their format.The dialogues have been held roughly every two months and attended by both Executive Vice-President Dombrovskis and Commissioner for Economy Gentiloni.The dialogues take MEPs' questions in slots of two or three interventions.MEPs intervene on behalf of their political group in decreasing size order-the European People's Party (EPP) group has the first slot, the Socialists & Democrats (S&D) the second, etc.After each slot, the questions are addressed by Dombrovskis and then Gentiloni until all political groups have taken the floor.Towards the end of the meeting, there is some-times time left for a "catch the eye" procedure, where other MEPs, who had not been originally selected by their groups, ask questions.
The dialogues have an average duration of about 1 hr 30 min, with a minimum of 10 MEPs (in the sixth dialogue on 2 May 2022) and a maximum of 16 MEPs (in three cases) taking the floor.There has been significant variation in the number of questions asked to commissioners, ranging from 23 to 42 per dialogue.On average, each member taking the floor asked about two questions, although a few members asked four questions or more in their interventions.In total, our dataset includes 334 questions asked in 140 interventions by MEPs.
We start by assessing the strength of the accountability relationship between the EP and the Commission, focusing on the content of the questions asked by MEPs and the answers provided by commissioners.Then, we investigate the level to which the EP's oversight is directed.We analyse the strength and level of accountability both in the aggregate and over time.Finally, we move on to who takes the floor in the RRDs by analysing the number of interventions made by MEPs according to their party and country.

Strength of Accountability
Questions can be weaker or stronger instruments of accountability.Questions aiming to gather further information on the RRF, its functioning or its implementation are weak oversight instruments.Questions that compel the Commission to justify its actions publicly or require the Commission to change its behaviour based on a negative judgment of its recent deeds are instead stronger accountability tools.Of course, there is also the possibility that MEPs are not using the dialogues as an instrument to hold the Commission accountable but rather as a forum to broadly discuss future policies and explore the Commission's position on the matter.For this reason, the questions are sometimes unrelated to the declared scope of the dialogues.
Looking at Figure 1, it is evident that the largest share (about 50%) of questions is to request information.Also, there is a significant percentage requesting policy views or discussing irrelevant matters.Together, questions requesting justification or a change of behaviour and/or policy amount to slightly more than 10% only.All in all, in the dialogues, the commissioners are not faced with hard scrutiny by MEPs, as they are mainly required to provide additional information on the process set in motion by the RRF regulation-and modified by the REPowerEU regulation-or to express their views on future policy developments.Rarely do the MEPs confront the commissioners, asking them to defend (or even change) their past decisions and actions.
Moving to the answers provided by the commissioners, we can see that Dombrovskis and Gentiloni generally provide explicit replies.As Figure 2 shows, this is the case across all types of questions, and it is so even for questions that we deem irrelevant in the context of the dialogues.Of course, from an accountability perspective, justifying past behaviour or actions and (not) endorsing a policy change is more significant.Here, the commissioners also do well, although, on the latter, the share of intermediate replies equals that of explicit replies.Sometimes, commissioners remain vague, addressing the questions in general terms only; in other cases, they point to the legal competencies of other institutions or the national level.Across the dialogues held so far, the relationship between the EP and the Commission appears to be more cooperative (with MEPs wanting to know more about the process, the implementation, the views of the Commission, etc.) than antagonistic (with commissioners "grilled" by MEPs on the fault lines of the NRRPs and their implementation).

Level of Accountability
Art. 26 of the RRF regulation tasks the EP not only to hold the Commission to account ("EU level") but also to scrutinise the implementation of the RRF in and across the member states ("national" and "transnational" level).In this respect, Figure 3 shows that almost 80% of all questions target the EU level, with MEPs focusing mainly on the role of the Commission, the interpretation of the rules, or other EU-level institutional or policy issues.Overall, MEPs seem to behave as European-rather than national-representatives, as only about 10% of the questions are asked by MEPs targeting their own country.The same share of questions focuses on the transnational level, showing that MEPs do not often question the Commission on issues of implementation in other member states and when they do, the focus is mainly placed on Hungary and Poland.
To be better appreciated, these findings should be placed in a comparative context.The literature on parliamentary questions has shown that the share of questions with a national and/or regional focus pertaining to the MEPs' own member state varies between about one-fifth and one-quarter of all such interventions (Brack & Costa, 2019, pp. 236-237;see Chiru, 2022, pp. 278-280, for a recent review).The substantively smaller national focus of the questions in the RRDs could be interpreted as (early) evidence that MEPs are using this tool to mainly scrutinise EU policymaking, leaving the inquiry over national-level implementation to their domestic counterparts.However, the growing saliency of the implementation process might change these trends in the future.In this respect, some insight might come from analysing the dialogues over time.

Development of the Dialogues Over Time
As we analysed dialogues that were held across two years (from May 2021 to April 2023), we can expect various factors to drive the variation of questions over time: the changing political context (i.e., the war in Ukraine), the different implementation phases of the RRF, changes in the RRF architecture (i.e., the REPowerEU chapters), and the experience gained by the MEPs and the commissioners using the instrument.
First, we expected that the type of scrutiny would change.In the early dialogues, questions should be more on gathering information, given the novelty of the RRF and the need to "find out" how the new procedure would work.In later dialogues, MEPs should be better equipped to demand explanations and even ask the Commission to change its policies or behaviour.As Figure 4 shows, this is not the case-there seems to be no clear pattern.In September 2022, for instance, MEPs had been asking as many questions requesting information as questions requesting justification or change.In the subsequent dialogue held in November 2022, MEPs barely asked questions requesting justification/change.All in all, the share of questions requesting information remains over 70% across our range, with one exception.
Second, we expected MEPs to focus more on the national and the transnational levels over time, as the disbursement of the money and the implementation of the plans was progressing across the EU member countries.Figure 5 partly confirms this expectation.Over time, MEPs targeted the EU level less and the member state level more, with both the "transnational" and the "national" levels becoming more important.
Third, experience should also matter for commissioners answering MEPs' questions.expected to know better both the RRF instrument and the format of the dialogue with the EP.This expectation is partly supported by Figure 6, showing that the share of explicit replies peaked in 2023, while that of non-replies diminished year after year.

Who Takes the Floor?
The BUDG-ECON joint Committee is composedincluding substitutes-of 52 members of the EPP, 40 of the S&D, 30 of Renew Europe, down to 10 MEPs for The Left.The distribution of speaking time by political group should proportionally replicate the numerical strengths of the political groups, given that the D'Hondt method is used to allocate speaking slots among the groups.However, Figure 7 shows that this pattern is only partly followed.the joint committee), the most active MEPs are not the members of the EPP, but rather those belonging to the S&D.The Greens/European Free Alliance (Greens/EFA) members are also more active than those of the larger Renew Europe group.The European Conservatives and Reformists (ECR) and The Left are overrepresented, while the Identity and Democracy group is slightly underrepresented.Unsurprisingly, non-attached (NA) members take the floor the least.The rather disproportional distribution of speaking time among the groups suggests that some groups-specifically, those beyond the "core" of the EP party system (i.e., the Greens/EFA, the ECR, and The Left)-are keener to intervene in the dialogues.
While decisions on the allocation of speaking time to MEPs are taken within the political groups-and more in-depth research is needed to understand their internal allocation process-interesting information can be gathered by comparing the national composition of the joint committee with the nationality of MEPs taking the floor in the dialogues.Based on the size of the national  This imbalance may reflect the redistributive nature of the RRF, whose impact is very significant on the economy of some member countries (mainly in Southern and Eastern Europe).However, it is rather negligible in others (particularly in Northern Europe).Of course, this finding might cast doubt on the dialogues as an instrument of EU accountability.Given the institutional innovation brought in by the RRF-with the mutualisation of debt to finance the instrument with EU money and, consequently, reforms and investments in the member states (Fabbrini, 2022)-it is surprising that the MEPs from countries with more "hawkish" attitudes towards debt mutualisation and EU fiscal "solidarity" have intervened so little in the dialogue with commissioners.MEPs from the member countries mostly impacted by the EU financial resources are, instead, the ones in charge of the dialogue with the Commission.
Nevertheless, while this is true, it is not the whole truth.Another remarkable feature has to do with the fact that there is a relatively restricted number of "expert" MEPs who play a key role in the dialogues.Those members who are either chairs or vice-chairs of the BUDG and ECON Committees, and those MEPs who have been rapporteurs or shadow rapporteurs of the regulations on the RRF and REPowerEU, made 87 of the 140 interventions in the RRDs (about 62%).Incidentally, MEPs from Spain, Portugal, and Romania took the lion's share of (shadow) rapporteurships on legislation on post-pandemic recovery.In this sense, both nationality and expertise "matter" to explain who takes the floor, with the experts drawn particularly from Southern and Eastern member countries.

Conclusions
Through an in-depth analysis of the 10 RRDs between May 2021 and April 2023, the article has empirically examined the strength and type of accountability exercised by the EP vis-à-vis the Commission within the legal framework set by the RRF.Similar to the economic and monetary dialogues, we find that the achieved accountability was overall limited.Our analysis suggests that in a (post-)crisis context dominated by executive actors, the scrutiny of post-pandemic recovery by the EP has not been particularly strong.If, more generally, parliaments play a secondary and mainly reactive role in the governance of crises, the story of the EP is not exceptional.It is the executive actors at both the EU (the Council and the Commission) and the member states level (the national governments) which are in charge of the NRRPs, assessing developments, authorising payments, and implementing projects and reforms.The EP is regularly informed about the progress, and the RRDs require the commissioners to report before the joint BUDG-ECON Committee.However, not only has the EP not been granted the possibility to address its questions to the Council or national ministers, but the dialogues provide, at best, a "light" form of accountability vis-à-vis the Commission.Only one question out of 10 requested justification or change on the part of the Commission, while most questions merely asked for additional information, followed by requests for policy views or opinions on irrelevant issues.On the other hand, the EP's questions are, by and large, directed to the EU level.They focussed on the rules, implementation, and problems of the RRF as a whole-as an EU project.In the period covered by the analysis, MEPs rarely asked questions about their own country or scrutinised what happened in other member states.This latter aspect has, so far at least, been a missed opportunity for MEPs to cross-check how member states spend EU money.As the success of the NGEU ultimately depends on the effective implementation of the NRRPs (the domestic reforms/milestones as well as the thousands of national and sub-national projects) the EP should also place delays, improper use of funding, and cases of corruption within the EU member countries under the spotlight.
We have also observed that there is little "learning" in the process: the differences between the earlier and later dialogues are limited, except for the increase in the explicitness of the commissioners' answers.This might be explained by the persistence of an information asymmetry between the commissioners (who oversee the implementation of the NRRPs in close dialogue and cooperation with the member states) and the MEPs.The political uncertainty surrounding the RRF and its subsequent developments, including the new challenges brought about by the energy crisis and the Russian invasion of Ukraine, likely contributed to maintaining this gap.
From a policy perspective, limited institutional changes could help strengthen the oversight of the RRF (cf.Chang & Hodson, 2019;Bressanelli, 2022).The joint BUDG-ECON Committee could promote a stronger collaboration among MEPs, with a greater provision of information about the implementation of the NRRPs in all member countries and the possibility of follow-up questions for unsatisfactory answers by the commissioners.Other executive actors, such as the president of the Council of the EU or national ministers, could also be invited to attend the dialogues.However, this change would be more challenging, requiring an amendment to the RRF regulation.
The article has also explored, in a preliminary fashion, the profiles of the MEPs who take the floor.The picture that emerges is very interesting: the most active MEPs are from Southern and Eastern Europe (Italy being a notable outlier), while parliamentarians from Northern Europe engage very little (Germany being an exception).At the same time, we find that the majority of the most active MEPs in the RRDs hold key positions in the EP committees and/or were involved in the negotiation of the RRF and/or REPowerEU regulations as (shadow) rapporteurs.It could be that these MEPs were selected to speak in the dialogues because of their specialised knowledge of the issues at hand, and they might have been considered in a better position to inspect and challenge the commissioners.There also seems to be a correlation between being an "expert" of NGEU and certain nationalities or geographical areas.Future research should disentangle and analyse such aspects in more depth.

Figure 1 .
Figure 1.The type of questions asked by the MEPs in the BUDG-ECON joint Committee.

Figure 2 .
Figure 2. MEPs' questions in the BUDG-ECON joint Committee and commissioners' answers.

Figure 5 .Figure 6 .
Figure 5.The level of MEPs' questions in the BUDG-ECON joint Committee over time.

Figure 7 .
Figure 7. Political group membership in the BUDG-ECON joint Committee and interventions in the RRDs.

Figure 8 .
Figure 8. Nationality of MEPs in the BUDG-ECON joint Committee and activity in the RRDs.
MEPs' type of questions in the BUDG-ECON joint Committee over time.