Reaching the poorest and hungry groups of the population, including those who might be left out of the Millennium Development Goals, involves input from policymakers at the central and local levels of government. While there has been considerable focus on appropriate targeting mechanisms to reach the poor, attention as to which level of government should be involved, as well as the interactions among levels of government in reaching the poor, is more recent. From a policy perspective, it is important to examine the instruments available at each level of government in order to meet the needs of the poorest. If the responsibility for these groups is seen primarily as that of the central government, then direct federal/central government programs— effectively targeted but building on local information—come into focus. If the primary responsibility is local, the policy focus shifts to own-source revenues for financing the expenditures and for greater local accountability, together with a modicum of equalization transfers so that all local governments have similar capacities to provide for the poorest.
Designing central programs to reach the poorest may be difficult without local information. This is because the central government, particularly in large countries such as China and Mexico, lacks the ability to precisely define marginal groups or households that may not benefit from more general growth and prosperity. Yet local officials may not share central government objectives, and they may prefer to divert central funds to meet objectives of higher value to them. This policy dilemma illustrates difficulties with “overlapping” responsibilities between different levels of government and in designing effective special purpose programs, financed centrally but implemented by local governments.
The constraints of designing effective, centrally determined special purpose programs in developing countries are legendary. To some extent these are similar to the problems of designing foreign assistance strategies that effectively reach the poorest target groups in the recipient countries: country elites may not share the altruistic objectives of donors. The issue is to design policies that build on local information yet minimize local incentives to divert resources from the target groups. Similar issues arise in the context of foreign assistance designed to reach the poorest groups of the population.
Local governments with the most limited resources and tax bases may also have the greatest requirements for supporting the poorest groups and individuals. Thus, providing for these groups in the poorer localities may require central-government assistance, even though the central government may also have concerns for poorer groups living in the relatively well-to-do localities. Urban poverty, even in the richer areas, is becoming an increasing problem in addition to more traditional rural poverty. Reaching these groups may not be easy and could generate moral hazard difficulties as local officials attempt to minimize their financing of such activities and shift the burden to the central government. A continuation of pockets of extreme poverty in richer localities tends to reflect divergent preferences between central- and local-government officials. Local politicians and officials may not be particularly interested in such groups, and indigent people frequently do not have a significant political voice.
Relatively incomplete information is available to the central government either on poor and hungry people lacking family support or on subnational budgetary operations. This generates incentives for subnational governments to divert central funds allocated for the poorest. This policy brief examines the fiscal instruments available to different levels of government and how they interact to enhance the effectiveness of public policies for the poorest and hungry groups.
Addressing the leakages associated with central funding and local implementation requires that programs be designed to incorporate competition for resources—both across jurisdictions and over time. This also involves increasing the results orientation of budget processes and improving mechanisms for intertemporal feedback in the future allocation of resources.