Did tiny loans to the poor really end poverty?

For some borrowers, a little; for ending poverty, the evidence says much less than was hoped.

The story begins with a very small sum. Inspired by the Bangladesh famine of 1974, Muhammad Yunus lent a total of US$27 to 42 people, mostly bamboo-stool makers, as start-up money, without the burden of predatory interest. His Grameen Bank went on to provide small loans to the poor without collateral, relying on group lending and weekly instalments. It inspired similar projects in more than 64 countries, and in 2006 the bank and Yunus shared the Nobel Peace Prize. (The idea has older roots: microfinance can be traced to the 1720s Irish Loan Funds set up under Jonathan Swift.)

Proponents say microcredit reduces poverty through higher employment and incomes. The test came with randomised trials. In 2008 Abhijit Banerjee's team at MIT's Poverty Action Lab ran the first randomised evaluation of introducing microcredit in a new market, in slums in Hyderabad, India. Fifteen to 18 months after lending began, average monthly spending per person had not changed, though spending on durable goods rose. New businesses increased by one third, but were not very profitable.

A critical review of 58 papers covering 18 countries found no good evidence of a beneficent impact on the well-being of the poor, and noted that the greatest impacts came from the weakest studies. One summary puts it this way: microcredit has achieved much less than its proponents said, but its harms have not been as drastic as some critics argued. Critics point to debt traps in Andhra Pradesh, India, and to Banco Compartamos charging 86% per year in Mexico in 2007.