TOMS impact evaluation finds zero to negative effects in El Salvador

This is from the Economist:

The first of two studies found that TOMS was not wrecking local markets. On average, for every 20 pairs of shoes donated, people bought just one fewer pair locally—a statistically insignificant effect. The second study also found that the children liked the shoes. Some boys complained they were for “pregnant women” and some mothers griped that they didn’t have laces. But more than 90% of the children wore them.

Unfortunately, the academics failed to find much other good news. They found handing out the free shoes had no effect on overall shoelessness, shoe ownership (older shoes were presumably thrown away), general health, foot health or self-esteem. “We thought we might find at least something,” laments Bruce Wydick, one of the academics. “They were a welcome gift to the children…but they were not transformative.”

More worrying, whereas 66% of the children who were not given the shoes agreed that “others should provide for the needs of my family”, among those who were given the shoes the proportion rose to 79%. “It’s easier to stomach aid-dependency when it comes with tangible impacts,” says Mr Wydick.

For a litany of criticisms of TOMS before the study see here, here, and here. The original study is available here.

Also, would anyone ever think that donating shoes, or even mining hard hats, to rural Kentucky would be “transformative”?

Anyway, huge props to TOMS for daring to scientifically study the impact of their ill-advised in-kind aid initiative.

How to hack a Latin American election

Bloomberg has a fascinating longread on Andrés Sepúlveda, a man who claims that he rigged elections all over Latin America for almost a decade.

Here is an excerpt:

Sepúlveda’s career began in 2005, and his first jobs were small—mostly defacing campaign websites and breaking into opponents’ donor databases. Within a few years he was assembling teams that spied, stole, and smeared on behalf of presidential campaigns across Latin America. He wasn’t cheap, but his services were extensive. For $12,000 a month, a customer hired a crew that could hack smartphones, spoof and clone Web pages, and send mass e-mails and texts. The premium package, at $20,000 a month, also included a full range of digital interception, attack, decryption, and defense. The jobs were carefully laundered through layers of middlemen and consultants. Sepúlveda says many of the candidates he helped might not even have known about his role; he says he met only a few.

His teams worked on presidential elections in Nicaragua, Panama, Honduras, El Salvador, Colombia, Mexico, Costa Rica, Guatemala, and Venezuela. Campaigns mentioned in this story were contacted through former and current spokespeople; none but Mexico’s PRI and the campaign of Guatemala’s National Advancement Party would comment.

The whole thing, to borrow from Tyler Cowen’s vocabulary, is self-recommending. I am not a Latin Americanist and so cannot vouch for the veracity of the story. But it’s definitely a fantastic read.

And of course, as happens, a multi-country story of this nature is incomplete without casual stereotyping of hackers based on nationality:

For most jobs, Sepúlveda assembled a crew and operated out of rental homes and apartments in Bogotá. He had a rotating group of 7 to 15 hackers brought in from across Latin America, drawing on the various regions’ specialties. Brazilians, in his view, develop the best malware. Venezuelans and Ecuadoreans are superb at scanning systems and software for vulnerabilities. Argentines are mobile intercept artists. Mexicans are masterly hackers in general but talk too much. Sepúlveda used them only in emergencies.