The NGO survey is a very good effort.

I wish they had published more of their methodology, and included a copy of the questionnaire used.

Did you notice that, of the 32 provinces, they have data from only eight? It's not because those eight are viewed as representative, either. It's because those are the only eight provinces where it's relatively safe for those administering the questionnaires and conducting the interviews to travel around.

To protect respondents, NGO staff and programmes, the survey was carried out in secure areas only. Many areas of the country were already "off limits" for many NGOs.

Stallard, Dawn. "Afghan Opinions on Rights and Responsibilities". Pakistan [no city]: The Army Press/Human Rights Research and Advocacy Consortium, 2003. p. 5, footnote 2.

Doctors Without Frontiers pulled out of Afghanistan after decades in the region, including during the Soviet occupation and the civil war between the war lords and the Taliban. Security must be very bad.

On the other hand, I heard an optimistic report on NPR (you heard me ;) ) this afternoon about the upcoming presidential election in Afghanistan. They have no shortage of candidates. Let's see if there is actual voting in all 32 provinces, or if there's voting in eight provinces and polling places blowing up in the rest.