Rethinking Poverty

IMF News, August 19, 2016

Source details

There are "a billion poor people in the world," but they are largely invisible to most policymakers and citizens.
Collecting data is "very expensive," and without data "we don’t see them" and "we need to actually focus on the problem."
Finding: intuitive understanding of what is easy or difficult for poor people is unreliable; rigorous data collection and focused study are necessary.

Position: Banerjee believes poverty traps exist ("do I believe it is true? Yes, I believe it’s true") but notes evidence is mixed and hard to establish definitively.
Evidence cited: interventions that help people today and leave them richer many years later suggest a possible trap, because beneficiaries do not always fall back into poverty.
Implication: targeted identification of the specific constraint for each person (e.g., money vs confidence) would improve efficiency, but achieving that level of sophistication may be difficult; "one shouldn’t let the perfect be the enemy of the good."

Clarification: "Aid" as commonly understood (foreign aid) is a small part of social support; most social support is provided by local governments and local charities.
Assertion: "Real resources count. Better quality education, more assets for poor people—those things are real money, and real money needs to be spent."
Critique of microcredit: the promise that micro-credit is merely "lubrication" without substantial resource transfer is characterized as "mostly false."

Diagnosis: governments often "give [money] away to the wrong people"; a major problem is subsidies that disproportionately benefit the rich.
Historical note: problematic subsidies are a "sad leftover from the 1970s and 1980s" when subsidies were politically preferred.
Recommendation: the first priority is to "stop giving the subsidies to the rich," freeing up resources that "you would have lots of money for the poor" if subsidies were removed—acknowledging this is politically difficult.
Preferential policy instrument: direct transfers are "more efficient than subsidies," especially because power and water subsidies are "extraordinarily distortionary," environmentally harmful, and inequitable—benefiting "people who have large power-using machines" more than the poor.

Progress: there is "a fast-growing body of evidence" that clarifies "what are really stupid things to do and what actually works"; this evidence is increasingly penetrating policy levels.
Institutional barrier: translating evidence into practice is hampered by bureaucrats who "have their own entirely unwarranted opinions"; reaching the government bureaucrat who decides on programs will "take a while."
Outlook: the shift toward evidence-based anti-poverty policy "has left the station" but changing on-the-ground decision-making is described as "a long battle."

IMF News: Rethinking Poverty, August 19, 2016.