مشاجرات بدهی های مستقل: پایگاه داده در دولت اجباری طی بحران بدهی

This paper measures “debt disputes” between governments and foreign private creditors in periods of sovereign debt crises. We construct an index of government coerciveness, consisting of 9 objective sub-indicators. Each of these sub-indicators captures unilateral government actions imposed on foreign banks and bondholders. The results provide the first systematic account of debt crises that goes beyond a binary categorization of default versus non-default. Overall, government behavior and rhetoric show a strong variability, ranging from highly confrontational to very smooth crisis resolution processes. In a preliminary analysis on the determinants of coercive behavior, we find political institutions to be significant, while economic and financial factors play a lesser role. These results open up an agenda for future research.

Sovereign debt crises are usually regarded as binary events: A government is either in default or it is not. This paper develops a more refined approach to analyze debt crises and debt renegotiations. We argue that the binary categorization for default versus non-default is overly simplistic, as it ignores the large variation in crisis resolution policies and related negotiation patterns. Our aim is to measure the wide range of debtor policies once a country has entered a default or debt restructuring process. A comparison of the recent crises cases in Uruguay 2003 and Argentina 2001-2005 illustrates our main point. Argentina’s government halted all of its debt payments for several years, refused to negotiate with creditors and enforced a unilateral debt exchange in 2005. In contrast, Uruguay avoided any missed payments, engaged in close creditor talks and arranged a voluntary debt exchange within just three months. We argue that these cases are not the same and, in principle, should not be treated as the same in empirical research. To overcome the missing procedural knowledge on debt crises, we develop an index of government coerciveness, capturing confrontational debtor policies vis-à-vis private external creditors in times of debt distress. To construct this index we draw partly on criteria suggested by the IMF, 1999 and IMF, 2002 and the Institute of International Finance (IIF, 2006). Specifically, we draw on the IMF’s “Policy of Lending into Arrears”, which made any emergency financing conditional on “good faith” efforts in resolving a debt crisis. Good faith debtor behavior, according to the IMF, includes a transparent debt workout process, early and continuous dialogues with creditors, and data sharing. A similar code of conduct was set up in the IIF’s “Principles of Fair Debt Restructuring”, signed by over 30 countries and supported by the G7, the G20, the World Bank and the IMF. The IIF defines restructuring processes as fair, if debtor governments closely cooperate with creditors, if they adhere to information sharing, avoid unjustified capital controls, and if they resume partial or full debt service payments as soon as conditions allow. Building on theses and further contributions we develop an index with 9 objective sub-indicators. Each sub-indicator captures unilateral government actions that governments impose on foreign banks and bondholders. They can be categorized into measures of “payment behavior” (4 sub-indicators) and measures of “negotiation behavior” (5 sub-indicators). The final index is additive, with a minimum value of 1 (low coerciveness) and a maximum value of 10 (very high coerciveness) and is measured for each debt crisis year. The resulting yearly database starts in 1980 and covers 251 crisis-year episodes in 31 developing countries that defaulted on sovereign debt. We also code criteria for each restructuring separately, covering a sample of 101 sovereign debt restructurings with private external creditors. With the coerciveness index and database we provide the first quantitative account on debt crises beyond a simple default dummy. More generally, we are the first to code debt renegotiation processes and related disputes between governments and private international creditors for a large sample of financial crises. The index also improves on earlier attempts to categorize debt crises or debtor coerciveness by Cline (2004) or Roubini (2004) among others. A main advantage is that our coding approach is reproducible and comprehensive in scope. We conducted a systematic evaluation of more than 20,000 pages of articles from the financial press, of main reference books and data sources on debt crises, and of numerous policy reports. Furthermore, the measurement approach can be applied to different eras of debt restructurings. The criteria chosen are general enough to compare debtor coerciveness across debt crises and restructurings of the last three decades, despite the shift from bank to bond financing and a changing role of actors such as the IMF. In this paper we provide a detailed account of our measurement approach. Overall, the results show an impressive variance in government negotiation behavior and rhetoric towards private creditors, ranging from very confrontational behavior to very smooth crisis resolution processes. We portray main stylized facts and discuss what can be learned from the categorization of government behavior. We also conduct a first explorative analysis on the determinants of debtor coerciveness. The regression results indicate that political and institutional factors are important for the degree of debt disputes, while many economic or financial factors are not. More specifically, we find most of the “rules of thumb” driving sovereign risk (Manasse and Roubini, 2009) to be insignificant predictors of coerciveness. This opens up an agenda for future research. The structure of the paper is as follows: Section 2 discusses the related literature and previous attempts to categorize debt crises. Section 3 presents the “Index of Government Coerciveness” and each of its 9 sub-indicators from a conceptual point of view. Section 4 outlines the coding procedure and the datasets that resulted from it. Section 5 briefly presents some descriptive statistics and the main stylized facts revealed by the data. Section 6 provides first evidence on the determinants of coercive behavior. Finally, Section 7 concludes and provides an overview of research questions that could be tackled with the new datasets.

This article provides the first comprehensive and systematic account of government behavior during debt crises that goes beyond a binary measure of default versus non-default. We assess how sovereigns resolve debt crises and which coercive actions they impose on their private international creditors during debt renegotiations. Overall, we find a strong variability in crisis resolution patterns across space and time. The sub-indicators are general enough to accommodate changes in restructuring mechanisms, instruments, actors and third party policies such as those of the IMF. They may also be suitable to systematize historical debt crises of the 19th and early 20th century and to evaluate future instances of sovereign default. A number of key insights emerge from the data: First, on average, governments behaved somewhat more cooperatively during the 1980s debt crises than during the Brady and Post-Brady eras. The volatility of the index has increased since 1998, with the Argentinean bond restructuring of 2001–2005 as a notable outlier of particularly coercive behavior. Second, there seem to be serial patterns of coerciveness. Countries with governments that adapted a conflictive stance in debt crises of the 1980s also tended to show unilateral government behavior in the 1990s and in more recent restructuring cases. Third, there are important differences regarding the type of debt restructured. On average, recent negotiations to restructure bank debt and domestic bonds were of more conflictive nature than foreign bond restructuring processes. In a last step, we also conduct a first analysis on the determinants of government coerciveness. We find most economic and financial variables to matter little, while political and institutional variables play a significant role. These results motivate future research on the political economy of debt crises and crisis resolution. More generally, the dataset may be used to tackle a whole set of unanswered research questions, even beyond the arena of sovereign debt. Why are financial distress situations resolved in such different ways? What determines debtor–creditor relations in times of crises? And which political and economic factors drive government policies and rhetoric towards financial market participants? Furthermore, there may be negative consequences of unilateral debtor policies, which have neither been analyzed nor understood. An analysis of the reputational costs of aggressive debt policies could provide new insights to the academic literature and debate. Lastly, policymakers and practitioners may use the index and data as a benchmark to assess future instances of default and debt renegotiations.