Interview
Do unconditional cash transfers cause economic growth in low-income countries? | Paul Niehaus (2023)
- Cash transfers are expected to generate broad economic ripple effects in interconnected systems, with a general equilibrium multiplier effect estimated at 2.5 in rural Kenya, compared to a 1.5 to 1.9 range for U.S. stimulus spending, driven primarily by the utilization of underutilized labor and capital rather than fixed capital expansion.
- GiveDirectly's approach challenges the status quo by asserting that cash is the default resource allocation unless a confident alternative exists, though exceptions may occur where top-down approaches outperform recipient-led spending in specific contexts.
- Empirical evidence is predicted to show no significant negative outcomes such as reduced labor supply or increased spending on alcohol and tobacco, while housing improvements like concrete floors and metal roofs are anticipated as common uses that improve nutrition and reduce disease risk by enabling clean rainwater collection.
- Investment returns on assets purchased with transfers are expected to range from 20% to 50% annually, with potential for structural change to concentrate in rural retail and small-scale manufacturing rather than large-scale urban migration, although statistical detection of treatment effects may fade over time as control groups benefit from economic spillovers.
- In the Rwanda "Hugoku Dekore" comparison, cash transfers are forecast to raise monthly income by 99 inverse hyperbolic sign points against 28 points for the workforce program, leading to the expectation that cash will likely be a lower bar for employment goals but superior for overall consumption and well-being outcomes.
- USAID benchmarking studies are expected to be revolutionary by testing conventional programming against cash, potentially revealing that cash transfers outperform targeted employment programs in most metrics except for specific employment generation, which is characterized as a "private good" problem difficult for outsiders to influence.
- Scalability and reach are central to the outlook, with projections that GiveDirectly could assist every person in extreme poverty at enormous scale, absorbing more funding than other charities due to implementation efficiency and relevance, while noting that deworming is a regional solution whereas cash is applicable globally.
- Risk management expectations include a fraud and theft cost of roughly 10 to 20 basis points (approx. 1% of funds) being a reasonable cost of doing business, with specific policy changes prohibiting field staff from registering SIM cards following a nearly $1 million fraud incident in the DRC.
- The DRC fraud case, representing less than 1% of 2022 deliveries, is not expected to shake confidence in the mission, with continued operations planned for fragile and conflict-affected regions where the extreme poor are increasingly concentrated.
- A 12-year Universal Basic Income (UBI) study in Kenya aims to test how future income expectations affect risk-taking and investment, predicting significantly larger economic effects from the long-term arm compared to a short-term 2-year arm, even with identical total cash amounts, while labor supply shifts from wage employment to self-employment are anticipated without significant changes in total hours worked.
- Recipients are expected to prefer lump-sum payments over monthly installments due to barriers to saving and the need for capital for infrequent large purchases, despite counter-arguments regarding budgeting for lean seasons, with data expected to show no negative impacts on the subjective well-being of non-recipients.
- Price pressure from widespread cash injections is predicted to remain very small (fractions of a percentage point) in well-integrated regions with underutilized capacity, though effectiveness may be limited in dense urban areas where capacity is already fully utilized.
- The organization expects to continue refining cost-effectiveness calculations by eventually incorporating indirect community impacts, noting that current models may appear more favorable than they are, while acknowledging that GDP growth does not always equate to well-being if it results from longer working hours and reduced leisure.
- Regional variations in effectiveness are anticipated but remain under-tested due to a lack of designed comparisons, and waiting for interventions to definitively beat cash is viewed as a tragedy that would leave hundreds of billions of development dollars idle.
- GiveDirectly's role is defined as providing a challenge function and a hurdle for other charities, working complementarily with GiveWell, while the speaker notes that measuring total value by converting diverse outcomes into a single number is too complicated, preferring to treat the dollar as a numeraire for comparison.