Wednesday, December 24, 2008

Hajj pilgrim and randomized experiments

Randomized experiments seems to keep moving the frontier. From micro credit to deworming to teacher's absenteeism to corruption. Recently, a study on the impact of hajj pilgrim and tolerance behavior was conducted using the method, by Kennedy School's Ashim Khwaja and his colleagues.

This is how they did the randomization (in which I am interested more):

For the study, three economists interviewed 1,600 Pakistanis, half of whom had been on the hajj in 2006 and half of whom had applied for visas to go but were rejected. Respondents answered 200 questions in face-to-face interviews that lasted hours. The interviews took place five to eight months after the pilgrims came home.

And this is the result:

Muslims who undertake the hajj "return with more positive views towards people from other countries," are more likely to say "that people of different religions are equal," and are twice as likely as other religious Muslims to condemn Osama bin Laden, the study found.

Looks like economists can now have job opportunities in the Ministry of Religion...


  1. This comment has been removed by the author.

  2. Could it be, though, that the people who were rejected during their visa applications were actually denied a visa because of reasons correlated with those xenophobic/hostile/fanatic characteristics?

    That seems a more likely explanation than "the hajj completely transforms its pilgrims into a tolerant, open-minded, and anti-Osama people". Keep in mind, there was no "before" interview, only an "after", so we can't really track changes in individual behavior.

    Although, I must admit, it would be much better if it were the latter explanation rather than the former.

  3. throughout the paper, the authors seem to draw a lot of insights from social psychology literature, suggesting *contact* with people from other countries during the hajj as the main mechanism behind the pilgrims' change of views.

    yet, at the same time, they argue that the hajj itself is responsible for the more positive characteristics.

    surely the two are nowhere near equatable. so this, i find, is rather preplexing.

  4. Very smart point Fik.

    There could be a self selection bias in the population going for hajj.

    Those wanting to go already has some calling and behavioral propensity so the process of hajj itself might be enhancing rather than changing.

    would be great to know the degree and depth of interaction during hajj and how it correlated with the result.

    Might also interesting to compare between haji abidin (finance by state/other people) and self finance with hajj cost as percentage of income as additional variable.

  5. Taufik/Berly: yes, it is a possible threat to validity, indeed. But does the hajj visa process really look or filter such behavior or indications (like the process to get a US visa, for example)? I think quota allocated to each country matters more for someone to obtain a hajj visa. Can anyone who has been to the pilgrim or applied for the visa confirm this? Also, remember that this study is specific to Pakistanis hajj pilgrimers.

    Tirta: my impression was they attributed the behavior change to the whole hajj experience. Anyway, this is a journalist report of a study so I'll check with the original paper. But, can you tell me how did you get such impression? Maybe I overlooked the article.

  6. @ape: it's just my overall impression after skimming the paper. in any case, follow-up studies are definitely in order (e.g. difference-in-differences type of studies tracking change of beliefs and attitudes as a result of exposure to members of outgroups would help clarify the present issue).