Table 5

The Role of Assortative Preferences

 Subject’s gender
Female
(1)		Male
(2)		Female
(3)		Male
(4)
Age
    Man is between zero and five years older0.039
(0.003)***		0.061
(0.004)***		  
    Man is five years older or more  −0.054
(0.005)***		−0.040
(0.006)***
    Woman is older  −0.022
(0.005)***		−0.081
(0.006)***
Education
    Similar education level0.017
(0.003)***		0.022
(0.004)***		  
    Man is more educated  −0.022
(0.086)		−0.088
(0.083)
    Woman is more educated  −0.000
(0.086)		0.051
(0.083)
Occupation
    Similar occupation0.007
(0.005)		0.006
(0.006)		  
    Both students  −0.010
(0.100)		−0.241
(0.125)*
    Both self-employed  0.106
(0.131)		−0.309
(0.163)*
    Both manual  −0.012
(0.023)		0.001
(0.028)
    Both skilled / nonmanual  0.001
(0.007)		0.002
(0.008)
    Both professional/ manual  0.010
(0.007)		0.012
(0.009)
Smoking
    Similar smoking status0.025
(0.012)**		0.014
(0.014)		  
    Both smoke  0.045
(0.089)		0.068
(0.088)
    Both do not smoke  0.003
(0.086)		−0.040
(0.084)
Height
    Man is between zero and seven centimeters taller0.006
(0.005)		0.028
(0.006)***		  
    Man is more than seven centimeters taller  0.021
(0.007)***		−0.009
(0.008)
    Woman is taller  −0.032
(0.006)***		−0.046
(0.008)***
Weight
    Similar BMI0.000
(0.007)		0.012
(0.009)		  
    Woman overweight and man not overweight  0.109
(0.121)		0.047
(0.087)
    Man overweight and woman not overweight  −0.119
(0.125)		0.014
(0.071)
Observations46,06546,06546,06546,065
R20.320.340.320.18

  • Note: Estimates are obtained from linear probability models including subject and partner fixed effects. The similarity variables are defined as follows: Education: both no degree or both with university degree, Occupation: both students, both self-employed, both skilled nonmanual, both manual, both professional/managerial; Smoking status: Both smoking, both not smoking; BMI: Both underweight, both normal weight, both overweight. Observations are at the subject-partner meeting level.

  • ↵* significant at 5 percent;

  • ↵** significant at 1 percent.