Building Social Protection and Labor Systems : Concepts and Operational Implications

Abstract

This paper presents a framework for designing and implementing social protection and labor (SP&L) systems in middle and low income countries. Although the term 'system' is used to describe a country's set of social protection programs, these tend to operate independently with little or no coordination even when they have the same policy objective and target similar population groups. The paper argues that enhancing coordination across SP&L policies, programs, and administrative tools has the potential to enhance both individual program performance as well as the overall provision of social protection across programs. The first part of the paper discusses the characteristics of well?designed social protection systems. It also points to the gains and some of the risks - of moving toward systems, including: (i) more effective risk management in crisis and non?crisis periods; (ii) improved financial sustainability; (iii) more equitable redistribution; (iv) economies of scale in administration; and (v) better incentives. The second part discusses issues related to design and implementation based on country studies for Brazil, Chile, India, Niger, Romania, and Vietnam. It suggests three levels of engagement to support the design of SP&L systems: (a) at the policy level, defining how different instruments (e.g., savings, risk pooling, redistribution) interact, and coordinating financing mechanisms and institutional arrangements; (b) at the program level, improving the design of individual programs and creating synergies with other programs within and across social protection functions; and (c) at the administrative level, setting up basic 'nuts and bolts' tools that can work across programs, such as beneficiary identification and registry, payment mechanisms, and management information systems. The last part of the paper outlines some of the implications of a systems vision for the World Bank's social protection and labor practice.