Carrying out public opinion polls in Afghanistan is never going to be easy. Areas that are remote at the best of times are now inaccessible and dangerous.

But a representative poll of the Afghan people is exactly what the Afghan Centre for Socio-Economic and Opinion Research (ACSOR) has once more attempted – with surprising success. You may not have read much about it, as the UK media largely ignored its findings and even the BBC – which helped to pay for it – did not make much of it.
 
The ACSOR conducted 1,532 face-to-face interviews between the 11 and 23 December, each one lasting about half an hour and covering topics ranging from progress (both since the fall of the Taliban and over the past year) to support for the foreign armies and perceptions of last year's elections.
 
To ensure a representative sample, 194 sampling points were distributed between the 34 provinces of Afghanistan, proportionally to population size. Then, within each province, sampling points were allocated randomly to villages or neighbourhoods. (More details on methodology are available here.) 
 
Of the 194 finally selected, 21 proved impossible to access, generally for security reasons, so other villages or settlements within the same districts were randomly chosen as replacements. Not ideal, certainly, but the ACSOR should be applauded for the lengths it has gone to in order to be as representative of Afghanistan as possible.
 
This not good enough for Ryan Grim of the Huffington Post who spoke to four experts: a researcher, a former US foreign service officer who quit last year in protest at the US Administration’s policy, and two journalists, and whose story is headlined: “Experts on Afghanistan Doubt Survey on Foreign Occupation: Results are Impossible”.  He argues firstly that the sample cannot be representative as the areas that were inaccessible will have been more likely to be pro-Taliban. This may be true, but it is clearly acknowledged by the survey team, and 90 per cent of the areas first chosen were indeed sampled.
 
Secondly, he argues that people will have lied to the survey, either to attempt to please the US, or to try to benefit their village. But the actual responses to the questions do not appear to substantiate this point, particularly as there is evidently no restraint in criticising the West. The complaint that it would be impossible to get a 50-50 split between men and women respondents is also unfair, given that women were purposely targeted by women interviewers to ensure such a split. In the end it wasn’t quite achieved, but the ratio of 730 women to 804 men is not so far out.
 
The greatest potential flaw in the sampling highlighted by these critics is the charge that interviewers were free to choose the doors they knocked on, therefore may have have asked friends and family, creating an unrepresentative sample.
 
Gary Langer, director of polling at ABC News, who commissioned the poll with the BBC and Germany's ARD, has rebutted these claims, arguing: "We’ve spent many hours vetting its methodology, attending interviewer training, reviewing sampling plans and parsing the data”. It’s impossible to know how truly random the  interviewers were - they may sometimes have chosen a friendlier face or a house on the safer side of a street - but given the circumstances, these findings are bout as good as reasonably possible.
 
As for the findings, they are a mixed bag, as one would expect in a country with such huge problems. On the positive side, 70 per cent of people believe that Afghanistan is going in the right direction, up a remarkable 30 percentage points from the 40 per cent recorded in the same survey last year. This same pattern seems to be reflected in responses to questions about conditions of life. (See chart below, noting that the survey was not done in 2008 and rights of women was not included in the 2005 survey.)
 

                   Percentage of respondents who rated conditions good

 

 
Support for the Taliban, however, increased to 10 per cent (from 4 per cent in 2006) and foreign jihadi fighters were even more popular, with 17 per cent of respondents at least “somewhat supporting” them, up from 11 per cent last year. Furthermore, a majority (43 per cent) of those surveyed said they thought an “Islamic state, where religious authorities have final say in all political matters” would be the best form of government for Afghanistan, compared to the 32 per cent in favour of democracy and the 23 per cent who preferred a dictatorship.

Additional ambiguities are apparent in the Afghans’ assessment of the elections held last year. While 75 per cent of people described themselves as satisfied with the outcome of the presidential elections, with almost half of those “very satisfied”, 59 per cent believe there was fraud in the counting of votes and 57 per cent think that voting was not conducted honestly.

President Karzai's approval ratings are also surprising from a Western perspective. More than half of respondents (55 per cent) view him “very favourably” and another 28 per cent “'somewhat favourably” This increase seems to reflect the changing views of the Afghans rather than faults in the survey, as the proportion who rate his work as “good” or “excellent” this year has risen by 20 percentage points compared to last. In previous years they have been highly critical, which argues against the idea that they may have been fearful of speaking badly about the government.
 
Perhaps most important for Western governments are the opinions of the Afgan people about the foreign militaries: 68 per cent support the presence US forces, 61 per cent the 'surge' of 30,000 more troops and 62 per cent the presence of NATO forces. Most encouragingly, only 8 per cent think attacks on NATO forces can be justified, down from 25 per cent last year. Yet only 7 per cent are 'very confident' in the foreign troops' ability to provide security and stability, in comparison to 43 per cent for the Afghan national Army.
 
Whatever the criticisms and the difficulties, these poll results are interesting. But you could be forgiven for having missed this entire story, so under-reported was it within the British media. Perhaps the generally more positive outlook given by these figures ran contrary to the image most editors want to portray of Afghanistan.  One conclusion that one might easily draw is that the media  do not care for the Afghan people's opinions, only for the safety of British soldiers, even when those soldiers are in the country specifically so the Afghan people have the chance to voice their opinions.
 
The most remarkable conclusions emerged when respondents were asked how conditions now compared to those before the invasion. The results were overwhelmingly positive, despite the fact that Afghanistan remains a battlefield (see chart below). A large majority (83 per cent) believes the 2001 invasion was good, up from 69 per cent last year. You didn’t read that in the British press, apart from a decent piece in The Guardian, and passing mentions in The Times and The Independent. 
 
Why not? Maybe, like the Huffington Post, they all discounted it as impossible. Hard as it is to hear the people of Afghanistan in the confusion of war, it is sad that so many are unwilling even to listen. 
 

              How do these conditions compare to before the fall of the Taliban?