Friends or family? Revisiting the effects of high school popularity on adult earnings

Abstract

Recent evidence has suggested links between high school popularity and wages during mid-life using the Wisconsin Longitudinal Study. This article revisits this question by first replicating the results using an alternative dataset that is very similar in structure. Similar to previous results, the Add Health baseline effects suggest that an additional high school friendship nomination is linked to a 2% increase in earnings around age 30. However, leveraging the unique structure of the Add Health shows that sibling comparisons eliminate any associations between popularity and earnings. The findings suggest that families, rather than friends, may be the cause of the association.

Keywords:

• social networks
• popularity
• earnings
• sibling fixed effects

JEL Classification:

• J31
• J24
• J20

Notes

¹ Similarly, Fletcher (2011) shows that the impacts of breastfeeding on later outcomes often disappear when sibling comparisons are employed.

² There are two linked data collection activities in Add Health. There was an original (‘Wave 0’) in-school survey of 90 000 children that ascertained friendship nominations and basic demographic information. Secondly, there are the four longitudinal ‘in-home’ surveys that track 20 000 children. Approximately 75% of the 20 000 children in the in-home sample were also in the in-school sample. 

³ It is also important to note that 80% of the sample has a sibling; however, in order to be sampled in Add Health, the sibling needed to be in one of the 120 schools and in grades 7–12 in 1994/95. 

⁴ The midpoint of each interval is used in the analysis. The intervals include: $0, <$5000, $5000–9999, 10 000–14 999, 15 000–19 999, 20 000–24 999, 25 000–29 999, 30 000–39 999, 40 000–49 999, 50 000–74 999, 75 000–99 999, 100 000–149 999, 150 000 or more.

⁵ The Add Health Picture Vocabulary Test (AHPVT) is a computerized, abridged version of the Peabody Picture Vocabulary Test-Revised (PPVT-R). The AHPVT is a test of hearing vocabulary, designed for persons aged 2½ to 40 years old who can see and hear reasonably well and who understand standard English to some degree. Each test included a set of practice, or pretest items, followed by a series of test items arranged in order of increasing difficulty. The respondent was asked to listen to the word spoken by the interviewer and to select the picture on the plate that he or she believed best illustrated the meaning of the stimulus word. Once the response was entered into the computer, the program indicated the next plate to use in the test. In addition, the computer program determined test results automatically. These test results were not made available to the interviewer or to the respondent. The test scores are standardized by age. Some psychologists interpret PVT scores as a measure of verbal IQ. Information on the test is provided online at http://www.cpc.unc.edu/projects/addhealth/files/w3cdbk/w3doc.zip