Mexico International Migration

INTERNATIONAL
1

Hi!

I was looking at the questionnaire for Mexico Population census and housing 2010. There are a few questions in this questionnaire that I cannot find in the dataset regarding international migration:

  1. Date of emigration Q8
  2. Current country of residence Q11
  3. Date of return Q12
  4. Current resident condition Q13
2

These variables come from a supplemental data file. Specifically, the details for the Mexico 2010 migration supplementary file are available here. This file contains information on household members who moved abroad in the previous 5 years and may have returned.

3

Thanks Jeff!

I am just getting confused with the variable in “migid” from the supplemental data and “intmig2” from the main dataset.

Correct me if I am wrong:

  1. Supplemental: “migid” tells us in the last 5 years how many international migrants are there.
    Main data: “intmig2” tells us prior to June 2005 how many international migrants were there.

  2. Another point is that I can only know when a person returned from June 2005- 2010 if the person went abroad after june 2005 from the supplemental data. However, I will not be able to know when a person returned if he/she went prior to June 2005.

4

The key difference between the migration data in the supplemental file and the migration data in the core census file, for the Mexico 2010 sample, is that respondents in the core census file must be present in the household during the time of the census. This is not the case for those included in the supplemental migration file, who could be currently abroad or have already returned. Your second point is correct about the limitation of the supplemental migration file. For feasibility and quality purposes data collection efforts, like this one, need to set boundaries on the information they gather in some way.

5

Thanks, Jeff!
This helps.

6

Following up:

I saw that the supplemental data is updated.
Is there any way to link the persons currently living in the household from the supplemental data to the main census? I do not seem to find a unique identifier for each person to link the two datasets. I can link the households but not the individuals.

Thanks,
Avinandan

7

The data need to be manipulated a bit before you can complete a person-level merge due to the fact that there are a large number of cases where MIGID=99 (people who have migrated and not returned home) that can not be merged at the person-level to the original Census data. Follow these directions to complete a multi-step merge:

  1. Split the migration supplement into two person-level datasets: one for migrants with MIGID=99 and another for migrants with MIGID!=99 (everyone else)
  2. Create a household-level Census dataset from your extract using PERNUM=1
  3. Merge the MIGID!=99 (everyone else) dataset with the original Census data at the person-level using SAMPLE + SERIAL + PERNUM (use MIGID for the migration data)
  4. Merge the MIGID=99 dataset with the household-level Census data using SAMPLE + SERIAL
  5. Append the two datasets together
8

Thanks, Grace!
I tried the above-mentioned way but I don’t think MIGID=99 means people who have migrated and not returned home. I see 9476 individuals with the value of MIGID=99 have a year of return (yearret). Have they returned or is there a problem?

9

A returned person may not have necessarily come back to the same household. The code MIGID=99 is a “Not in Universe (NIU)” code for migrants who have returned from their migration but not to the same household (migrants who have returned to the same household receive a MIGID). A more precise indicator for migrant residence is RESIDM (current residence status of migrant), question 13 from the migration questionnaire which indicates whether the person is “currently living in this dwelling”. As the table below shows, most cases where MIGID=99 are also RESIDM=“No” (the person is not currently living in this dwelling) or RESIDM=9 (Not in Universe (NIU)). There are 6,144 cases where MIGID=99 and RESIDM=“Yes”; these are interpreted to be unknown cases due to missing information. The code RESIDM=9 is mostly NIU but it includes some unknown cases. There are about 120k cases that are consistently in the no response category for MONTHRET, YEARRET, and RESIDM. These are people not living in Mexico according to COUNTRYR.

You can read more about these variables in the Migration Supplement on the INEGI website. I hope this helps!

image