Several studies have provided insights into the correlation of outcomes from World Bank DPF operations (as measured by overall outcome ratings from ICRRs) and characteristics of those operations.8
However, given the complexity of the results chain underpinning most policy reforms, these studies do not establish causality. 

Smets and Knack9
explored how the World Bank’s policy-based lending has influenced the quality of economic policy, given the focus of DPF on policy reforms as immediate and intermediate objectives. Their analysis focused on the early part of the broad results chain of aid outlined by Bourguignon and Sundberg,10
from aid to policy making and policies, but it did not address the subsequent link between policies and country outcomes. To this end, the authors used elements of the World Bank’s Country Policy and Institutional Assessment (CPIA) ratings as the dependent variable and as a proxy for the quality of economic policy. Although past studies found only limited association between DPF operations and the macroeconomic stability component of the CPIA, the authors examined those links and focused specifically on macroeconomic and fiscal issues, in particular macroeconomic stability as measured by the relevant CPIA indexes. They did not find an association between DPF operations and macroeconomic stability. However, they did not address reverse causality, i.e., the possibility that the DPF operation was selected because of stable macroeconomic conditions. This is important because the World Bank requires macroeconomic stability at the time of approval. 

Moll, Geli, and Saavedra11
empirically tested whether elements of operations design (for example, prior actions, results frameworks, and task team leader skills and professional affiliation) influenced the success of World Bank DPF, as measured by IEG-validated overall outcomes for 2004–2013. They controlled for income per capita, quality of macroeconomic and governance-related policies, and force majeure as measured by natural disasters. They found that the line of sight between the policy reform supported and the results framework was critical for success, while the skills of the task team leader and a professional affiliation with the Poverty Reduction and Economic Management network of the World Bank was also associated with greater DPF success. 

Bogetić and Smets12
found that the World Bank’s policy lending was significantly and positively correlated with the quality of social policies and institutions. They also found that loan conditions related to social protection and environmental sustainability were associated with better social policies and institutions more than those related to the equity of public resource use, and with health and education. These findings reinforce the idea that the type and quality of World Bank conditionality matters for the design and outcomes of policy-based lending.

• 8
  
  For a retrospective overview of Bank’s DPF, see World Bank 2015. World Bank Development Policy Financing Retrospective, Results and Sustainability. World Bank.
• 9
  
  Lodewijk Smets and Stephen Knack. 2014. World Bank Lending and the Quality of Economic Policy. World Bank Policy Research Paper. No. 6924.
• 10
  
  Francois Bourguignon and Mark Sundberg. 2007. Aid Effectiveness: Opening the Black Box. American Economic Review. Vol. 97, No. 2, pp. 328-21
• 11
  
  Peter Moll, Patricia Geli, and Pablo Saavedra. 2015. Correlates of Success in World Bank Development Policy Lending. World Bank Policy Research Paper. No. 7181.
• 12
  
  Željko Bogetić and Lodewijk Smets. 2017. Association of World Bank Policy Lending with Social Development Policies and Institutions. World Bank Policy Research Paper. No. 8263.