The supplemental .zip archive for this publication contains the following files:

1. STATA do-file to obtain the relevant variables from the original LIK data set
	name of file: 01_variables from modules.do
2. STATA do-file to identify couples and build the final data set to be used for analysis
	name of file: 02_build final dataset
3. STATA do-file to create random couples and calculate personality similarity indexes for them
	name of file: 03_generate random couples
4. STATA do-file to produce Table 2: Estimation results: Association of profile similarity index and type of marriage
	name of file: 04_main regression results
5. STATA do-file to produce Table 3: Estimation results: Association of personality trait similarities index and type of marriage, controlling for marriage duration
	name of file: 05_regression results separate personality traits



Below are instructions for reproducing the following descriptive finding: 
Becker, Charles M. and Steiner, Susan (2019). How marriages based on bride capture differ. Evidence from Kyrgyzstan. Demographic Research.

********************************************************************************************************************************** 

We use the Life in Kyrgyzstan (LIK) Study which can be accessed at the Data Set Repository at the IZA Institute of Labor Economics. 
https://datasets.iza.org/

In order to get access to the scientific use files proceed as follows:
    Register and download the contract form.
    Enter the requested information on the contract form, sign it, and send it to the International Data Service Center (IDSC) of IZA user support.
    After the application is approved by the IDSC of IZA the scientific use file will be provided as download for registered users.

Our analysis primarily used wave 2013 of the LIK. For some individuals and some variables, information had to be retrieved from the 2012 and 2011 waves. 


Create folders C:\Data\LiK2013, C:\Data\LiK2012, and C:\Data\LiK2011 to save the respective waves on your computer. Also create a folder C:\Replication to save all data files to produced with the above do-files. 



**** LIK 2013 variables used:
hhid = household identifier
pid = personal identifier

from data set named hh1a (household roster)
h102 = sex 
h103a = age 
h105 = ethnicity 
h108a = personal identifier of spouse or cohabitating partner

from data set named id2a (education)
i211 = highest education degree obtained

from data set named id2c (personality)
i244_1 to i244_21 = level of agreement to 21 items which form the basis for calculation of the Big Five personality traits

from dataset named id5d (women's background and fertility)
i519 = number of marriages
i520_1 = own age at first marriage
i520_2 = own age at second marriage
i521_1 = husband's age at first marriage
i521_2 = husband's age at second marriage
i522_1 = type of first marriage
i522_2 = type of second marriage

from data set named cc_hh (control card file)
soato = community code
psu = primary sampling unit
oblast = province

from data set named cc_ind (control card file; necessary to merge individuals across time)
hhid12 = household identifier in 2012
pid12 = individual identifier in 2012
hhid13 = household identifier in 2013
pid13 = individual identifier in 2013
hh_new = household new to survey in 2013
ind_new = individual new to survey in 2013
present = someone else present during individual interview



**** LIK 2012 variables used: 
hhid = household identifier (individuals cannot simply be tracked across time with this variable)
pid = personal identifier (individuals cannot simply be tracked across time with this variable)

from data set named id2 (education, health and personality)
i207 = highest education degree obtained
i225_01 to i225_21 = level of agreement to 21 items which form the basis for calculation of the Big Five personality traits

from data set named id5d (women's background and fertility)
i519 = number of marriages
i520_1 = own age at first marriage
i520_2 = own age at second marriage
i521_1 = husband's age at first marriage
i521_2 = husband's age at second marriage
i522_1 = type of first marriage
i522_2 = type of second marriage

from data set named cc_ind (control card file; necessary to merge individuals across time)
hhid11 pid11 hhid12 pid12 i_attr
**** LIK 2011 variables used: 
hhid = household identifier (individuals cannot simply be tracked across time with this variable)
pid = personal identifier (individuals cannot simply be tracked across time with this variable)

from data set named id2 (education)
i207 = highest education degree obtained

from data set named id5d (women's background and fertility)
i519 = number of marriages
i520_1 = own age at first marriage
i520_2 = own age at second marriage
i521_1 = husband's age at first marriage
i521_2 = husband's age at second marriage
i522_1 = type of first marriage
i522_2 = type of second marriage




STATA version 14.2 was used for the above do-files. ******************************************************************************************************************************
Contact: steiner@c4ed.org, date: 12/08/2019