FINANCIAL INFORMATION USE IN PARLIAMENTARY DEBATES IN A CHANGING CONTEXT

Several studies have analyzed how users perceive the accounting practice and financial information in the public sector. However, only a concise number of authors (e.g., van Helden et al., 2016; van Helden, 2016; Jorge et al., 2016b, 2019; Caruana & Farrugia, 2018; van Helden & Reichard, 2019; Giacomini, 2020; van Helden et al., 2021; Jethon & Reichard, 2022) specifically consider the perspectives of the politicians, as a group or compared to public sector managers, regarding the use and usefulness of financial information.

A few scholars (e.g., Guarini, 2016; Liguori & Steccolini, 2018; Jorge et al., 2019; Sedmihradská & Kučera, 2020; Faber & Budding, 2022; Raudla, 2022) uphold the relevance of the use of financial information by politicians in Parliaments, especially for political debates regarding budgets and governments assessment, but they also highlight that the selection and use of such information by politicians depend on different contexts, personal beliefs and preferences (e.g., Reichard & van Helden, 2016; van Helden & Reichard, 2019). Consequently, empirically analyzing the kind of financial information used, the frequency of use and intentions, underlying its use by politicians is a crucial addition to complement previous studies on this matter.

Accordingly, this study aims to analyze which type of financial information is used in parliamentary debates and the intentions for its use, considering a context of change (recovery of the economic and financial situation and a new accounting system, and a politically different legislature). Intentions behind the use of financial information are expected behaviors by the politicians when they resort to such information, according to their affiliation with parties of the government and of the opposition (Saliterer et al., 2019).

In this research, financial information is understood in a broad sense, comprising budgetary information (estimated and executed budget, according to budgetary legislation), financial accounting information (financial reporting according to accounting standards), and National Accounts outputs (government financial statistics – GFS). Also, it includes both cash- and accrual-based information.

The study relies on content analysis and sentiment analysis of the transcriptions of debates on financial matters in the Portuguese Parliament during two legislatures (2011-2019), embracing changes in economic and political conditions.

This research contributes to the existing (but not so abundant) literature on the use of financial information by politicians in several ways. It focuses particularly on the use of financial information by politicians in Parliament at the central level; most literature regards the use of non-financial performance (management) information and addresses the local government, whose politicians usually are not under the scrutiny of (central) Parliament (van Helden, 2016; Nogueira et al., 2021). It adds in the type of information considered, going beyond the information produced by the accounting system to include budgetary information and that of the National Accounts. It is also innovative in the methodology, resorting to sentiment analysis techniques, combined with content analysis of speeches, which gives more objectivity to the findings, contrary to the prevalent use of (perceptions gathered from) questionnaires and interviews (Nogueira et al., 2021). Finally, using the Portuguese case between 2011 and 2019, this approach looks for an evolutionary analysis, addressing a context of several institutional changes. This is also an important contribution, considering that literature essentially focus the use of financial information by politicians in a particular moment in time or situation, for example, in the approval of the budget or of the budget execution.

Henceforth the paper is organized as follows. The next section explains the context of the financial information used in the Portuguese Parliament. Then, the literature review is summarized, especially addressing the use of financial information by politicians and its intentions; research questions are derived at the end. It follows a section describing the methodological issues, and another one presenting the main findings. The paper finishes discussing the results and synthesizing the main conclusions.

Portugal is a unitary state and a Parliamentarian Republic. It became a member of the European Union in 1986 and consequently it must follow requirements in the EU treaties and other regulation. Worth to be mentioned is the Stability and Growth Pact and the Maastricht fiscal discipline criteria underlying the Euro currency, namely public deficit and debt limits as in the National Accounts. Portugal became under these restrictions from 2002. During the Euro joining process, a new Budget Framework Law was passed (Law 91/2001) to accomplish with the fiscal requirements. This law was revised several times, the most important changes happened from 2011 onwards. Firstly, as a consequence of the financial and economic crisis and of an austerity policy required by the Memorandum of Understanding (MoU) between the Portuguese government and the ‘Troika’ (the European Commission, the European Central Bank and the International Monetary Fund). Then, in 2013 and 2014 other revisions happened to accomplish with the requirements of EU stricter fiscal targets, namely accommodating the Directive 2011/85/EU, which culminated the so-called ‘Six-Pack’ of EU legislation regarding fiscal discipline. To implement these revised Budget Framework Law, the lenders observed,¹ among other issues, handicaps in the existent public sector accounting (PSA) system, requiring Portugal to move towards a system based on the International Public Sector Accounting Standards (IPSAS). At the same time, the aforementioned EU Directive (article 3) required member states to put in place PSA systems covering all sub-sectors of the general government and endorsing accrual accounting, to support convergence between governmental accounting and the National Accounts. In 2013 the European Commission published a document (European Commission, 2013) about the suitability of the IPSAS to the member states. Subsequently, Portugal approved in 2015 a new Budget Framework Law (Law 151/2015), keeping the main budgetary principles as before, but endorsing a new PSA system and improving budgetary control and reporting, hence accomplishing with the EU requirements of more transparent and better-quality public sector financial information. Accordingly, the System of Accounting Standards for Public Administrations – SNC-AP (Decree-law 192/2015) was approved too.

During this process of change in the budgetary and accounting frameworks, in political terms the Portuguese Parliament² passed two different periods. In 2011-2015 (XII Legislature) there were six political parties and a parliamentary majority of center-right wing; this was a period of economic and financial crisis and austerity policy. In 2015-2019 (XIII Legislature) there were seven political parties, and the Parliament changed to center left-wing majority, as a result of a parliamentary agreement between left-wing parties. In this latter period, Portugal recovered from the economic and financial crisis and started implementing the new Budget Framework Law and the SNC-AP.

Regarding financial information possible to be used in parliamentary debates, Table 1 summarizes the main types, the typology of political actors involved, and the related debates.

Politicians are key managers and ultimate decision-makers in the public sector. Furthermore, they have obligations to discharge accountability towards citizens regarding the financial resources entrusted to them and the provision of public services; politicians in lower levels of government are also accountable to upper levels. As citizens representatives, politicians in the parliament scrutinize and assess governments and public administrations overall. Considering these general purposes, several authors have discussed the use/non-use of financial information by politicians, at different levels of government and, fewer, in parliaments.

The literature has focused some relevant aspects: i) characteristics of financial information that make it suitable for being used (types of information and documents); ii) politicians’ perceptions of the relevance of public sector financial information and its actual use (politicians’ preferences and behavior); and iii) circumstances and purposes for using by politicians – what information is used and how it is used (van Helden, 2016; van Helden & Reichard, 2019).

Regarding the characteristics of financial information, the study by Buylen and Christiaens (2015) mainly shows that (accrual-based) financial accounting information is not frequently used by politicians in their debates concerning budget approvals due to its complexity. (Cash-based) budgetary information is considered more appropriate for evaluating governments because politicians, in general, find it easier to understand. Further studies have also highlighted politicians’ difficulties in understanding financial information and consequently in using it, leading to the preference for using non-financial information (e.g., Liguori et al., 2012; Poljašević et al., 2021).

Other studies argue that the lower use of financial accounting information by politicians does not necessarily arise from a lack of understanding but depends on the information available and the format in which it is presented – reports that are too long and have too many figures can intimidate even financial experts (Caruana & Farrugia, 2018; Jorge et al., 2016a, 2019; Sedmihradská & Kučera, 2020).

Politicians have political agendas, and so, training them in financial matters is not a solution to increase their use of financial reports. Cash-based versus accrual-based information is not an issue either; politicians’ claims for financial information in a more integrated format is more likely what would increase its use (Caruana & Farrugia, 2018; Jorge et al., 2016a, 2016b).

Often, politicians use accrual-based information as that coming from the National Accounts. They need to evaluate and decide about the Government’s finance indicators, especially those relevant to government assessment by Members of Parliament (Jorge et al., 2016a, 2016b). Particularly the use of this type of information is affected by external actors (e.g., EU and IMF) towards whom a country may have to report. This ‘logic of discipline’ (Roberts, 2010) implies use of information in parliaments to debate the acquiescence of the requirements by, and agreements with, those institutions. The discussion by politicians is subordinated to technical issues related to international commitments, which determine the financial information they may use.

Regarding the types of information used by politicians, the literature has not addressed sufficiently the context of national parliaments, where countries’ financial matters are approached in a broader way, and the budget, budgetary process and the accomplishment with international targets, assume a key role (Cangiano et al., 2013; von Hagen & Harden 1995). Subsequently this study puts RQ1 as:

About politicians’ perceptions and behavior, Sinervo and Haapala (2019) and Jorge et al. (2019) underline that financial information used by politicians not only depends on their expertise, but also on their willingness to be duly informed and advised; the authors refer to other individual characteristics too, such as their political experience.

Additionally, politicians’ needs and use of financial information varies according to their role inside the parliament and in the budget process (Jorge et al., 2019; Roudla, 2022; Saliterer et al., 2019). For example, it is likely that members of committees dealing with budget and financial issues use financial information differently and more intensively than those in other non-financial committees. Also, while parliamentarians are more oversight oriented, governing politicians are responsible for preparing the budget, needing and using financial information differently (Posner & Park, 2007). Faber and Budding (2022), in a study including members of the Dutch Parliament and of the Central Government, conclude that the use of financial accounting information is higher by the members of Government than by the members of Parliament. Also, they observe no difference between the use of financial accounting information by members of parliament belonging to Government parties or to the opposition, Finally, their study evidences that new political parties in Parliament, as the younger ones, tend to use more financial accounting information, as well as those members that assume a financial role in the party.

Lastly, it seems that the intensity of financial information used by politicians also depends on the gap between information supply (information characteristics) and demand (politicians’ needs), meaning that politicians tend to make selective use of financial information (Buylen & Christiaens, 2015; Grossi & Reichard, 2016; Ouda & Klischewski, 2019; van Helden et al., 2021; Faber & Budding, 2022), sometimes using it only symbolically (Giacomini et al., 2016; Liguori & Steccolini, 2018), and not as a function of its actual usefulness.

As to the circumstances and purposes for using financial information, while some authors sustain they may be affected by the political decision-making process (Guarini, 2016; Giacomini et al., 2016; van Helden et al., 2021), others argue that politicians hardly use financial information for decision-making. It is used essentially for accountability in parliaments (Caruana & Farrugia, 2018), namely for debates within the budget process and for the government’s assessment (Buylen & Christiaens, 2015; Giacomini, 2020; Raudla, 2022).

Van Helden (2016) points out that politicians’ actual use of financial accounting information is much lower than their appreciation or perceived usefulness of those items. Later, van Helden and Reichard (2019) emphasize that the actual use of financial information depends essentially on the intentions of use. The authors explain that, in some circumstances, politicians may be inclined to use financial accounting information to accommodate their political preferences. Considering specific intentions, politicians may show they use financial information to a much higher extent (illusory) than the actual use. In a recent study, van Helden et al. (2021) emphasize the relevance of using financial accounting information concerning controversial projects, exemplifying with two case studies. They conclude that both ex-ante and ex-post types of financial accounting information use is associated with actors’ behaviors (regardless whether they are politicians or not), as ammunition or to justify the projects choice, moving from a negative to a positive standpoint, depending on their financial empowerment on those projects. By analyzing politicians’ speeches in municipalities, Sinervo and Haapala (2019) also concluded that local politicians use financial information to legitimize their opinions and decisions concerning the municipality’s financial situation.

Guarini (2016) highlights a preference both in the production and in the interpretation of financial accounting information by newly-elected local politicians as ammunition to attack previous governments.

Giacomini et al. (2016) analyze the uses of financial accounting information by local politicians, considering specific contexts, such as: different situations where decisions occur and the level of political conflict deriving from different politicians’ goals (political competition); and at a policy formulation phase, depending on the policy cycle. They conclude that the degree of political competition and conflict influences financial accounting information and the types of uses. If the conflict level is low, financial accounting information is used to support decisions. In contrast, if the conflict level is high, it is used to answer questions in policy formulation and to attack/legitimize economic policies during the decision-making process. Giacomini (2020) also analyses the use of financial accounting information by local politicians in different stages of the policy decision-making process (ex-ante and ex-post), addressing both the types and the intensity of use. He concludes that financial accounting information, although with similar importance during the policy process, have different roles depending on the moment when it is used: rational and ammunition at the ex-ante decision, and a more evaluative and justifying role at the ex-post monitoring, in order to convince the council members. Jethon and Reichard (2022) reach similar conclusions concerning the role of budgetary information.

Within the branch of literature addressing the purposes for using financial information by politicians, Saliterer et al. (2019) categorized politicians’ intentions while using performance-based budgets information. Their study was based in the Austrian parliament, when discussing a new experience with performance budgets. The analysis considered the perspectives of members of the parliament associated to the government party, and of those in the opposition. Four main categories of intentions were derived from their analysis – politicians may use information to: legitimizing and supporting; improving and understanding; de-legitimizing and ammunition; and deflecting and attention diversion.

Regarding the circumstances and behavior of politicians while using financial information, the above literature, with a few exceptions, focus the local government. Therefore, it is relevant to continue exploring those issues at central level, namely in the parliamentary context. So, RQ2 is:

A topic interesting to explore concerns the effects of contextual changes in the use of financial information by politicians at central level. Particularly changes in the political composition of the legislature and in the economic and financial situation in the country. Such topic appears to have been ignored in the literature so far, except by Guarini (2016). This author evidences in the local government of Italy that, when there is a political alternation, politicians tend to make more use of financial information, namely when political competition intensifies, by the increasing the number of political parties represented. Consequently, the present study tries to go further in this theme establishing RQ3 as:

All in all, the present research aims to analyze what type of financial information is indeed used, and with what intentions, by politicians in the Portuguese Parliament, in the debates on financial matters, in a context of change. It fills a gap in the above literature, as no study so far addressed an evolutionary changing context.

This research is descriptive and follows an interpretative nature, based on quantitative and qualitative content analysis (Bardin, 2009) and sentiment analysis (Luo et al., 2013; Kharde & Sonawane, 2016). It resorts to a methodology of text analytics and text mining (Struhl, 2015) to drill over Parliament’s records. Specifically, this investigation uses the journals with transcriptions of the debates on financial matters (as in Table 1) that took place in plenary sessions of the Portuguese Parliament from 20^th June 2011 to 24^th October 2019,³ reflecting a longitudinal analysis.

“Sentiment analysis can be defined as a process that automates mining of attitudes, opinions, views and emotions from text, speech, tweets and database sources through Natural Language Processing… involves classifying opinions in text into categories like ‘positive’ or ‘negative’ or ‘neutral’” (Kharde & Sonawane, 2016, p.5). It aims to determine the attitude of a writer/speaker concerning some topic (Luo et al., 2013).

Accounting concepts and financial information terms were defined from PSA and financial management literature, considering the scope of financial information established for this research. Accordingly, these terms relate to budgetary, financial accounting and National Accounts perspectives.⁴ This dictionary of terms was then used to support the directed content analysis (Hsieh & Shannon, 2005) finally performed. The use of terms as data for content analysis in political discourse followed the pattern already used in other studies involving politicians (e.g., Laver et al., 2003; Buylen & Christiaens, 2015).

Like that used by Buylen and Christiaens (2015), the scoring technique counted these keywords over the years under analysis, making it possible to perceive differences in the number of references per term from one legislature to another.

One did not differentiate between the acquaintance of the speaker with financial matters (for example, whether he/she can be considered ‘user with expertise’, namely being a member of the Budget Permanent Committee; or ‘user without expertise’). Also, the type of information used and the intentions of its use in the debates was not cross-referenced to the interventions per individual political parties. However, in the analysis of the intentions of use, it was considered whether the politician was related to the Government and the government political party or the parties in the opposition.

This second analysis combined the initial list of keywords with another dictionary of terms related to the intentions of use (sentiments), within categories like those by Saliterer et al. (2019). The type of speaker (Government or opposition) was considered too. Accordingly, four main categories of use were considered: legitimizing and supporting, and improving and understanding (positive sentiments); de-legitimizing and ammunition, and deflecting and attention diversion (negative sentiments). The dictionary of terms within these categories derived from the researchers’ interpretation, resulting of the analysis of the citations previously selected.

This second stage excluded from the analysis some parts of the speeches initially selected, where it was not possible to identify neither the type of speaker nor the type of sentiment/intention, according to the dictionaries. Subsequently, text mining was applied to a lower number of politicians’ interventions in the debates. After selecting relationships by the software, there was a random check by the researchers for reassurance.

The results of the analysis of the types of financial information are firstly presented in order to map their use in the debates. Table 2 displays the frequency and descriptive statistics of terms used by politicians. Table 3 additionally provides a categorization of these terms, according to the dimensions assumed in the scope of financial information. A total of 5,306 parts of speeches were considered (2,471 from the XII Legislature and 2,835 from the XIII Legislature), extracted from the journals based on the initial dictionary of financial terms.

Subsequently, the results about the intentions of using that information are presented, applying sentiment analysis to the politicians’ speeches. A total of 1,989 talks were used (1,025 from the XII Legislature and 964 from the XIII Legislature), fewer than in the first part, considering that in some fractions of speeches initially extracted, one could neither identify the sentiment nor the speaker.

This paper started by analyzing what specific types of financial information politicians intervening in the Parliament use in the debates on financial issues. It further explored the intentions with which such information is used by politicians in government-related parties and in the opposition.

Regarding the first research question, the most mentioned keywords were budget-related. Terms within Budgetary Information, either ex-ante or ex-post, were those most frequently used in the debates.

Most terms associated with accrual-based financial accounting information were never referred. This somewhat defeatist attitude on the use of non-budgetary financial information by politicians in the Parliament happened despite the PSA reform meanwhile started with SNC-AP, which was never referred to in the politicians’ speeches.

These findings reinforce the conclusion of previous research (e.g. Buylen & Christiaens 2015; Jorge et al., 2016a, 2016b, 2019; Caruana & Farrugia, 2018; and Giacomini, 2020), but using different sources of evidence – registrations of politicians’ interventions in the debates.

Given that the budget is the basis for political negotiations between the Government and the other political parties in Parliament, it is not surprising that politicians demonstrated their preference for items associated with Budgetary Information in detriment of those related to Financial Accounting Information. As in Reichard and van Helden (2016), budgets and fiscal control have historical importance in Parliaments, being central in the debates of financial issues (Cangiano et al., 2013; von Hagen & Harden 1995). Some accrual-based information from the National Accounts, namely about Deficit and Debt, related to critical targets concerning convergence within the EU, was used too, even if to a much less extent, following the imposed ‘logic of discipline’ (Roberts, 2010) (see section above on the context of the financial information used in the Portuguese Parliament). However, politicians never mentioned the keyword accrual accounting. This was also evidenced by Jorge et al. (2016a, 2019).

As to the second research question, about with what intentions politicians use financial information in the parliamentarian debates, considering negative and positive perspectives, this study concludes that, in general, politicians use financial information to defend their political opinions and measures. Those who belong to government-related political parties use financial information for legitimizing and supporting (positive sentiments). Those in the opposition, use it predominantly as ammunition against the current Government, or even to divert attention from measures they had implemented before (negative sentiments); less frequently, they can also use financial information to support and explain policies implemented when they were in power.

This evidence corroborates the research of, for example, van Helden and Reichard (2019), Saliterer et al. (2019), as well as that of Faber and Budding (2022), who highlighted that the actual use of financial information is related to politicians’ intentions, who use financial information adapted to their political preferences.

The same broad conclusion was emphasized by Giacomini et al. (2016), especially concerning the level of political conflict and competition. When this level is higher, the use of financial information increases either to attack or to legitimize public policies.

This is also aligned with the findings of Guarini (2016) and of Sinervo and Haapala (2019), according to whom politicians use financial information to sustain (further explain and legitimize) their opinions regarding the financial situation of their institutions.

The findings related to the third research question show that the use of financial information by politicians in the Portuguese Parliament is influenced by the changing context, once two different legislatures were considered, embracing distinctive political compositions and a alteration in the economic and financial condition of the country.

Whereas the types of preferred financial information in the debates have not changed, with Budgetary Information continuing to be privileged, intensity and intentions of using financial information by politicians were affected by those changes, as Posner and Park (2007) also underlined. In the XIII Legislature, when the Parliament composition represented a political alternation, moving from a right- to left-wing majority as mentioned, and consequently politicians previously government-related passed to the opposition, all categories of intentions of using financial information were evident with minor differences. Also, during this period the level of political competition intensified, because more political parties came to the Parliament and politicians then in the opposition were previously in Government and vice versa (change in the political cycle). These findings show, as in Guarini’s (2016) study, that more political competition implies a more intensive use of financial information in the debates.

Regarding the intentions of use, the main conclusions highlight different intentions of using financial information as politicians belong to the government-related parties or to the opposition, going along with the findings of Saliterer et al. (2019). The changing of political cycle evidenced that the intentions of use increased either for ammunition or even to divert attention from measures implemented when politicians were previously in power (negative sentiments), because in the second legislature there were more parties in the opposition. Likewise, the uses with positive intentions (positive sentiments) were more frequent in the XIII Legislature by previous government-related speakers (that became opposition) to legitimize and supporting as well as for improving and understanding, in order to defend their policies.

Therefore, the overall findings of this study allow to underline that the change in both the political and the economic and financial condition in a country affects the intensity and the intentions of using financial information by politicians in the parliament.

In the period of analysis, namely during the XIII Legislature, a new IPSAS-based PSA system started to be implemented in Portugal. However, this fact did not seem to have affected politicians’ financial information preferences and intentions of use in the Parliament. Perhaps due precisely to the fact that this new system has never been brought to the debate, politicians may have not been aware of the ‘new’ type of information they could use. This point was highlighted by Jorge et al. (2019).

This paper brings valuable additions, specifically extending the literature on the use of financial information to politicians in parliamentary debates, aiming at identifying what information they need. For the first time an evolutionary changing context is considered, highlighting the effect on that use of a political change in the Parliament composition (politicians moving from government-related parties to the opposition). Despite based on the Portuguese reality, the study provides essential insights for other realities, particularly about the types of financial information used by politicians, and intentions of that use, in changing contexts, either in the political scene or in the fiscal situation of the jurisdictions. Overall, the paper represents an important step towards a theoretical framework about the use of financial information by politicians.

As main limitations one can point the way the dictionaries have been pre-defined and built, especially the terms used in the sentiment analysis. The final list could not be ultimately validated, given the difficulties of contacting with the politicians regarding past speeches. Therefore, the selected terms mainly resulted from the authors’ interpretation, undergoing a certain degree of subjectivity. Furthermore, the use of software, if, on the one hand, may have brought robustness and exemption to the analysis, on the other hand may also have made it too automatic and blind, despite the researchers’ intervention for reassurance (Sydserff and Weetman, 2002).

This research can continue exploring whether financial information use changes according to a specific political party (e.g., left-wing or right-wing) and politicians’ financial (non) expertise. It can also address the role of external actors (e.g., EU and IMF) in affecting the use of financial information in central parliaments.