Essays in political economy and central banking

The present work collects four empirical papers in the field of political economy. Three of them, namely the first, the second and the fourth, explore the field of central banking from a political economy perspective. In particular, they investigate the relationship between central bank governance and economic outcomes. The third paper looks at the impact of fake news on electoral behaviour, but uses a similar text analysis approach of chapter two. All four chapters are empirical contributions that apply different techniques of panel data econometrics. Moreover, all four papers introduce new databases that I constructed during my PhD on my own or jointly with my co-authors. The first paper (Chapter 1) explores the impact of supervisory governance on financial stability. Scholars have long believed the governance of banking supervision to affect financial stability. Although the literature has identified at length the pros and cons of having either a central bank or a separate agency responsible for microprudential banking supervision, the advantages of having this task shared by both institutions (shared supervision) have received considerably less attention. Using a new database on supervisory governance in 116 countries from 1970 to 2016, this chapter studies the performance of three models of supervision: (1) supervision by the central bank, (2) supervision by an agency and (3) supervision by both the central bank and an agency. It finds that supervisory governance per se does not significantly affect non-performing loans. However, it also finds that, where the risk of capture is high, shared supervision is associated with a significant reduction in non-performing loans. This is in line with the supervisory capture theory, whereby it is more costly to capture two supervisors rather than one. Overall, these results provide new evidence in support of the relevance of supervisory governance in hampering supervisory capture from the banking sector. The second paper (Chapter 2) investigates whether the parliamentary scrutiny of central banks are effective in holding independent central banks accountable. As central banks expanded their role during the global financial crisis, demand for public scrutiny of their actions increased. While this revived the debate on which statutory setting works best to hold central banks accountable, empirical evidence on how current frameworks work is scarce. The aim of this work is to fill this gap. We focus on the parliamentary hearings of central banks, which are the most diffused arrangements to hold central banks accountable. Using text analysis techniques on the transcripts of the hearings, this work detects their key topics and the sentiments during the debates between the Bank of England, the European Central Bank, the Federal Reserve and their respective parliaments since 1999. The analysis shows that, while the objectives of the central bank play a relevant role in determining the topic of discussion, higher rates of unemployment are associated with a decrease in the focus of the hearings on price stability. Second, the paper finds that sentiments in the hearings are more negative when there is higher uncertainty and when actual inflation is more distant from the central bank’s inflation aim. However, while the correlation between drifts from the inflation target and sentiments signal an effective scrutiny, the significance of this result is weaker than the one for uncertainty. Overall, these findings suggest that parliamentarians use hearings to scrutinise the performance of central banks in line with their objectives and economic developments, but also that uncertainty plays an important role in setting the tone of the debates. The third paper (Chapter 3) investigates the effect of fake news dissemination on voting behaviour. Based on the case of the 2018 Italian elections, it exploits the linguistic differences of the region of Trentino Alto-Adige/Südtirol as an exogenous source of variation in exposure to fake news. The presence in the region of Italian voters with similar characteristics but belonging to either the Italian or the German linguistic group assigns individuals into distinct filter bubbles, each differently exposed to political misinformation in Italian. Using the share of Italian speakers in each municipality as instrument, the paper looks at the impact of fake news dissemination on the growth of populist platforms. In order to measure populism, it constructs a novel index applying text mining techniques on the Facebook posts of each party and their leaders over the course of the electoral campaign for the 2013 and 2018 elections. The paper finds that fake news had a negligible effect on populist vote in the region during the 2018 Italian general elections. The fourth paper (Chapter 4) belongs to the same literature of the first paper, but looks at a different aspect of supervisory governance, namely the political independence of supervisors. Since the crisis financial regulators and supervisors have been given increased independence from political bodies. But there is no clear evidence of the benefits of regulatory and supervisory independence on the stability of the banking sector. This paper fills this gap, introducing a new index of regulatory and supervisory independence for 43 countries from 1999-2019, covering aspects of institutional, budgetary and regulatory independence. It combines this index with bank-level data to investigate the effect of independence on financial stability. It finds that reforms that bring greater regulatory and supervisory independence are associated with lower non-performing loans, and higher capital in banks’ balance sheets. In addition, it provides evidence that these improvements do not come at the cost of credit growth or bank efficiency. Overall, the results of this paper show that increasing the independence of regulators and supervisors is beneficial for financial stability. The four chapters are at different stages of development. Chapters 1 and 3 have been published as working papers. Chapter 2 has been resubmitted to the ECB Working Paper series after having been reviewed by an anonymous referee. The fourth chapter is still work in progress.