When Ratings Don't Matter: Market segmentation and sovereign risk in the age of index investment, with Ben Cormier and Patrick Shea

Under what conditions do credit rating agencies (CRAs) discipline sovereign borrowers? Existing research both under and overestimates CRA influence in international political economy. Scholars have underestimated CRA power by focusing mostly on final rating decisions, missing how reporting sentiment affects both investors and governments, even when ratings remain unchanged. But research has simultaneously overestimated CRA reach by assuming all emerging market sovereigns are equally subject to rating agency discipline. Using novel quantitative text measures of CRA report sentiment and new bond issuance data, we show that CRA disciplinary power operates unevenly across a bifurcated sovereign debt market. The rise of index investment has created stark segmentation: some states have stable access to vast pools of passive capital through index inclusion, while others do not. We show that CRA report sentiment significantly affects borrowing costs for non-indexed sovereigns but have no effect on indexed countries. CRAs thus retain disciplinary power over precisely those states with the least bargaining power in international finance, while indexed countries have effectively escaped their reach. Our findings reveal how market segmentation reinforces hierarchy among borrowing states. 

The distributional effects of privatisation in IMF programs, with Merih Angin 

Privatization of state-owned enterprises (SOEs) has been a defining feature of Inter- national Monetary Fund (IMF) programs since the 1980s. Privatization is amongst the most important and politically contentious type of IMF conditionality as it leads to major structural changes in the economy through the transfer of ownership and control. However, few studies to date have systematically investigated the distributive impact of privatization conditions of IMF programs. We theorize that privatization reduces the labor share of income by weakening labor’s bargaining power through two channels: reducing labor’s capacity to disrupt production, and reducing labor unions’ ability to organize. We test our theory using a mixed-methods approach. We first trace the underlying causal mechanism of our argument through two typical case studies of IMF sponsored privatization in Pakistan and Turkey. To test our theory in a large-N setting, we then employ regression analysis on all IMF programs from 1980 to 2015 and find that IMF privatization conditions have a negative effect on the national labor share of income. 

Currency hierarchies and development bank financing after economic globalisation

According to the conventional wisdom on economic globalisation, ‘deep’ integration, which involves external and domestic liberalisation, should be incompatible with state owned national development banks (NDBs) which are used to control domestic credit allocation. Although NDBs have declined since the 1980s, reports of this decline have been exaggerated, and mask significant variation in the responses of different countries to the onslaught of pressures to reign in statist institutions. This paper argues that the structure of the international economy makes it easier for developed countries to retain statist financial institutions due to strong and stable demand for their public securities as a result of their position in the currency hierarchy. This allows them to fund development banks on international markets and sidestep domestic redistributive conflicts, which arise when banks are funded on national budgets. This is not an option available to most developing countries with soft currencies, whose public securities are seen as risky. Developing countries can only retain these statist institutions if they resolve domestic distributional challenges. This is illustrated through comparative case studies of the post-80s evolution of NDBs in countries with varying levels of economic development: Germany, Brazil, South Africa, and Pakistan, using original archival material.