I have a dataset that looks at asylum rates for asylum-seekers (refugee status). The data is broken down by country of origin and the host state (where they apply for asylum. I have been reading up thoroughly on the types of missing data (e.g. MAR, MCAR, etc), but I'm terribly confused as it applies to my own data for a couple of reasons.

First, after I calculate the rates, some of the rates are missing because they neither received asylum, nor were rejected. In other words, they did not received a decision and are thus pending. Therefore, there are no rates.

Second, some of the data are missing because it is probably that no one from that country applied for asylum in the host state in year t.

Would these be cases of missing not at random?. Most of the missing data examples seem to revolve around surveys, and I understand it when I read it. I am just having a hard time applying it to my own scenario.

Edited: the unit of analysis is dyad-year (host state and origin state in year T). The independent variables look at domestic explanations from the host state. So, for example, do far-Right parties in power affect asylum-rates? Do do differences in the major religion of the host state and the state of origin (where asylum-seekers originate) affect asylum-rates? I also tried to pool as much as I could with the DV in order to have sufficient amount of observations to lag the dependent variable. I originally wanted to use a GEE model, but read that it gives an unbiased explanation if the missing data are MCAR.

  • $\begingroup$ Is your unit of analysis the person or the country? $\endgroup$ – rolando2 Mar 16 '17 at 0:02
  • $\begingroup$ Hi, it's country. Country-year. $\endgroup$ – EJ16 Mar 16 '17 at 0:05
  • $\begingroup$ OK, I recommend you make your question text fully consistent with that. Your "First" paragraph sounds like it's describing data at the person level--individual experiences. Beyond that, I suspect a really good answer to your MNAR question would require familiarity with all the variables in your analysis. So you may have some additional explaining to do.... :-) $\endgroup$ – rolando2 Mar 16 '17 at 0:11
  • $\begingroup$ Editing now, to further lessen the confusion. thanks for the reply! $\endgroup$ – EJ16 Mar 16 '17 at 0:12

The first is a case of right-censoring. Survival analysis deals with these kinds of problems.

The second is a kind of selection bias, that you might or might not want to control for. Perhaps there are ways to control for it via instrumental variables such as distance between the countries, etc

| cite | improve this answer | |
  • $\begingroup$ I didn't think about a survival model. You mean like a Cox proportional? Also I thought about controlling for distance, but there are lot of migrants that immigrate to places that are not border states, such as in W. Europe. $\endgroup$ – EJ16 Mar 15 '17 at 23:46
  • 1
    $\begingroup$ Yes exactly. But then I guess you would need a "time since applying for asylum"-metric. I see. Just trying to be creative :) Maybe there are more relevant things guiding the process, such as size of previous asylum groups from that country etc. $\endgroup$ – salient Mar 15 '17 at 23:50
  • $\begingroup$ Some of the numbers really look random. For example, some individuals from country X will only go to country Y once in a twelve year period. Or not go in year T1 but then later in T3. Some of it is obvious, such as conflict. Others are maybe not fleeing conflict but for other reasons. This has given me a headache for some time now, so I decided to throw a life line out here on stackexchange. $\endgroup$ – EJ16 Mar 15 '17 at 23:56
  • $\begingroup$ If we take a step back, what is your research question? $\endgroup$ – salient Mar 15 '17 at 23:57
  • $\begingroup$ Very simply, it looks at what are some domestic explanations that affect asylum-granting. So, for example, do far-Right parties in power in the host state really affect asylum-granting? If the host state has a different majority religion compared to those applying for asylum, does that affect asylum-granting? etc. The asylum-rates are the DV. $\endgroup$ – EJ16 Mar 16 '17 at 0:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.