Policies as Determinants for Economic Performance

Abstract

This dissertation studies topics that have substantial policy implications for the public sector, centering around designing, assessing, and implementing policies for improving economic performances at the macro level. The first chapter first conducts a comprehensive review of the recent development of IMF lending, including an early wave of literature on examining the determinants of IMF lending programs, and the second wave of literature on assessing the impact of IMF lending programs (and the attached conditionality). Based on the comprehensive review and the observed increased reliance of IMF decision-making on the nowcasts data, I identify the nowcasting of IMF lending activities as an important topic for future research that would be particularly useful to aid in implementing IMF lending programs. A gap in the existing literature calls for a comprehensive study to identify the drivers of nowcast inefficiency for IMF lending, based on all available IMF loan programs data up to date. In light of the first chapter, the second chapter examines the accuracy of the IMF's assessments of crisis nowcasts that predicate program design. Analyzing an unprecedented 602 IMF loan programs from 1992 to 2019, we contradict the popular notion that IMF forecasts are generally optimistic. By disentangling the structure of the nowcast bias, we find the IMF systematically overestimates low-growth recoveries for Low-Income Countries' (LICs) GDPs while underestimating high-growth recoveries. Our unusually large sample allows us to document that Non-LICs nowcasts exhibit no statistically significant optimistic/pessimistic bias. We isolate the sources of inefficiencies in IMF nowcasts, including: (i) program objectives, (ii) program conditionality, (iii) geographic regions, (iv) global crises, and (v) geopolitics (elections, conflicts, disasters). In addition, we show that shorter nowcast horizons do not improve accuracy, and that GDP growth nowcasts improved substantially since 2013. Inflation nowcasts continue to struggle with efficiency as recently as 2018. The third chapter studies the topic of macroprudential policy. It examines empirically the role of macroprudential policy in addressing the effects of external shocks on financial stability. The baseline dynamic panel regressions show that an appreciation of the local exchange rate is associated with a subsequent increase in the domestic credit gap, while a prior tightening of macroprudential policies dampens this effect. These results are strong for small open economies, and robust when we explicitly account for endogeneity of macroprudential policy and changes in exchange rates. We also examine a feedback effect where strong domestic credit pulls in additional cross-border funding, potentially further increasing systemic risk, and find that targeted capital controls can alleviate this effect.