Guest post by Jeff Mosenkis of Innovations for Poverty Action.

Image credit: 3ie

• Academic types imagining a more #RealisticIndianaJones got it trending on twitter.
• 3ie has updated their comprehensive database of published development impact evaluations and asked “Is impact evaluation still on the rise?” Looking at annual numbers of publications through the first 3 quarters of 2015, they may be leveling off.
• A pretty harrowing account has come out of Juba from July 11th, when South Sudanese troops broke into a compound popular with foreign journalists and aid workers. They singled out Americans as an example, and raped, executed, and beat the people there. The U.N. peacekeeping force less than a mile away  ignored calls for help (as did the American embassy).
  • This comes after the U.N. peacekeeping force failed to protect a refugee camp in February from South Sudanese forces, resulting in 30 deaths and 123 injured.
  • And in July the peacekeepers did nothing as South Sundanese troops raped dozens of women nearby.

Ultimately, there seem to be few incentives for peacekeeping troops to actually keep peace, and few punishments for not doing anything.

• Despite continuing dangers there, a large number of Indian businessmen have chosen to stay, refusing planes sent by the Indian government to take them out. Their interesting story here. (h/t Rachel Strohm).
• Paper: Trying to replicate nearly all political science studies on comparative and international political economy from two journals, with a different way of dealing with missing data (multiple imputation instead of listwise deletion), makes key results disappear in nearly half the studies. (h/t Ryan Briggs)
• How Honduras became a little bit safer.

And winner for scariest Pokémon picture:

Playing #PokemonGo in #Cambodia #Laos #Vietnam? Beware of #landmines or anything that looks like an old bomb! #UXO pic.twitter.com/BDJnlMrefI

— Political-Military Affairs, US Dept of State (@StateDeptPM) August 15, 2016