Subscribe to E-Bulletin Donate to CEGA

Indirectly Measuring Political Attitudes in Pakistan

Development Challenge

In the 2012 fiscal year, the United States spent 48 billion dollars on military and humanitarian foreign aid.[1]  A significant portion of this money went to the Middle East, where the US has had a substantial military and humanitarian presence.  Since the source of aid can affect citizen participation in aid-funded programs, understanding public opinion is paramount to programmatic success.  Public opinion is difficult to measure in low resource settings, however, particularly when citizens feel their views will not be socially accepted.  This study employs an offer experiment, a unique method of soliciting revealed preferences, to measure anti-American attitudes in Pakistan.


This experiment was implemented simultaneously in three cities in Pakistan: Peshawar, Islamabad, and Dera Ghazi Khan (DGK).  Because regions within Pakistan may have been affected differently by US action in Afghanistan, the researchers systematically chose these cities to capture a range of perspectives. Peshawar was directly affected by the invasion, while Islamabad and DGK have experienced a sharp influx of refugees due to the unrest. 

The ethnic makeup of each city was integral to its inclusion in the study.  Peshawar and Islamabad both have large Pashtun populations, while DGK has a large Balochi population.  Pashtuns are an ethnic group that has historically supported and participated in the Taliban.  The province of Balochistan was a hub of a large secessionist movement, and its capital, Quetta, is home to a large faction of the Taliban. Recruiters targeted study migrants from the Federally Administered Tribal Areas (FATA), a region with a large Pashtun population.  Researchers were most interested in measuring anti-American attitudes in these populations, as it is in these communities that such attitudes are of greatest importance.

Evaluation Strategy

The researchers contracted with local firms to recruit 1152 literate men, aged 18 to 35, for the experimental sample.  Pakistani research staff administered the study in order to avoid the potentially compromising presence of a foreigner. Lab sessions took place over the course of one week in July of 2013.  Researchers assigned lab session appointments randomly in groups of 24.  All participants received 300 Pakistani Ruppee (Rs) payment for their time (approximately three US dollars or one fifth of a day’s wage.) 

Participants were offered an additional payment (either 100 or 500 Rs, one or five USD, respectively) if they expressed gratitude toward the institution (either US or LUMS), publicly or privately.  Subject treatments were randomly assigned across three variables, forming a total of eight treatment groups with 144 subjects per group, as follows:

  LUMS Funding  LUMS Funding US Funding US Funding
Private Statement 100 Rs 500 Rs 100 Rs 500 Rs
Public Statement 100 Rs 500 Rs 100 Rs 500 Rs

Additionally, researchers asked participants direct questions about political preferences on Japanese and American governments and the aid they provide. 

Results and Policy Implications

Findings suggest that political opinion on US aid is moderated by social factors; a smaller portion of participants was willing to express publicly their rejection to US payment. Participants offered the US payment were 8.2 percentage points more likely refuse the payment privately than publicly.  In addition, respondents assigned to the private US statement were 16.8 percentage points more likely to refuse the payment than were participants in the private LUMS statement group. Differences between the public and private LUMS rejection rates were not significant; the difference between public and private rejection rates is therefore specific to the US case.

Photo Credit: US Embassy in Paksitan. Displaced Paksitani carries food supplies provided by the US government. 

[1] US Overseas Grants and Loans Report. (2014). USAID.