The design of anti-poverty transfers to the ultra-poor: Asset transfers or cash transfers?

A major challenge faced by government in low-income countries is how to reduce poverty in a sustainable way. Recent evidence suggests that the “Targeting the Ultrapoor” (TUP) program pioneered by BRAC in Bangladesh has proven very effective and portable across diverse low-income settings. As a consequence, up to 30 governments were piloting variants of TUP by the end of 2015.

The TUP program consists of a large asset transfer, typically livestock, combined with training. A key open question is whether beneficiaries could do better with an equivalent cash transfer. The answer could determine the expected cost of reducing extreme poverty: the administrative cost of TUP is sizeable so that a program that transfers the cash equivalent of the asset-skill transfer would cost a fraction of the TUP program per household.

• This project uses a large-scale randomized control trial (RCT) to compare the classic TUP design to a modified design where beneficiaries have the choice of cash instead of an asset-skills bundle.
• The study is taking place in the in the four poorest districts in Southern Punjab, where 10 million of Pakistan’s 40 million poorest households reside.
• The project is informing the design of poverty reduction programs in Pakistan. Implemented jointly with Pakistan Poverty Alleviation Fund (PPAF), the research team is also collaborating with other stakeholders in Pakistan, such as the Benazir Income Support Program (BISP) and the Punjab Planning and Development Department.
• For the RCT, villages were randomly assigned to one of three groups: 1) treatment group receiving a menu of assets-skills bundles; 2) treatment group receiving the same menu of asset-skills bundles, but with the additional option of receiving an unconditional cash transfer of the same value as the asset-skill bundles; and 3) a control group.

The aim of the project is to track household outcomes over the longer term to understand:

• Whether households have a revealed preference for cash transfers over asset-skill transfers;
• Do the returns to offering asset or asset-cash transfers differ?;
• If so, does this divergence in returns stem from the imperfect functioning of specific markets, or does it stem from household decision making processes differing by the form in which transfers are provided?; and
• What are the general equilibrium and distributional consequences on village economies of each transfer program?

All these questions feed into an accurate cost-benefit analysis of the programs. As stakeholders in Pakistan and elsewhere have expressed, this evidence will play an important role in wider policy discussions about how anti-poverty programs can be most efficiently designed.