Thread by @MatthewStaiger

***Updated paper***

How do connections in the labor market shape intergenerational mobility? I shed light on this question by studying one type of connection: jobs obtained at a parent's employer.

🧵 #EconTwitter
matthewstaiger.github.io/matthewstaiger…

I combine LEHD + Decennial Census to construct a dataset with information on parent-child and employer-employee linkages and show that it’s surprisingly common for people to work for a parent’s employer, with 29 percent of individuals doing so by age 30. [2/19]

Is this about parents using connections to provide access to jobs or something else? Maybe parents and children have similar skill sets and this is what leads them to work for the same firm. [3/19]

Looking at parents who begin new jobs, I find that individuals are 5x more likely to work for a firm that their parent recently joined compared to a firm that their parent will join in the near future. This points to the connections story. [4/19]

Furthermore, young workers in blue-collar industries are most likely to work for a parent’s employer, and these are the same industries where survey data suggests that the use of social contacts in job search is most common. These also tend to be high-paying industries… [5/19]

Main empirical challenge is to estimate the earnings consequences: of individuals who work for their parent's employer, how much more do they earn at their parent's employer relative to their next best option? [6/19]

I instrument for whether someone starts their career at their parent’s employer using the hiring rate at the parent’s employer at the time they enter the labor market. Empirical model includes fixed effects for parent’s employer and local labor market conditions. [7/19]

To highlight the identifying variation, I show initial earnings are strongly related to the contemporaneous hiring rate at the parent’s employer but unrelated to hiring rate at parent’s employer measured in earlier years. [8/19]

Initial earnings of the child are also unrelated to hiring conditions at different firms that are similar to the parent’s firm. In short, my empirical strategy exploits transitory and idiosyncratic shocks to job availability at the parent’s firm. [9/19]

Headline finding: working for a parent’s employer increases initial earnings by 17 log points (or about 19%). This is a large increase in earnings. For comparison, it’s about 1 s.d. of the inter-industry wage premium or ¼ the college premium. [10/19]

Earnings of young workers are 10x more strongly correlated with hiring conditions at parent’s current employer compared to parent's future employer. So any threats to identification must apply to parent’s current but not future employer, which rules out a lot of concerns. [11/19]

The earnings gains appear to be driven by the types of firms that parents provide access to. Working for a parent’s employer makes individuals 31pp less likely to work in the unskilled service sector and 33pp more likely to work in the blue-collar sector. [12/19]

Earnings gains are attributable to parents providing access to firms that pay all workers more (column 1 is a pretty interesting validation of the AKM model). These tend to be more productive firms that occupy a higher rung of the job ladder. [13/19]

Earnings gains of starting career at parent’s employer persist for at least 3 years, but magnitude declines over time, as other workers are able to catch up by climbing the job ladder. [14/19]

What are the implications for intergenerational mobility? This depends on whether individuals from high- or low-income backgrounds benefits more. Not actually obvious who we should expect to benefit the most. [15/19]

It turns out that individuals with high-income parents are significantly more likely to start their careers at the same firms as their parents. This is because high-income parents are more likely to be employed and to hold positions of authority within their firm. [16/]

Individuals with high-income parents also experience larger earnings gains from working for a parent’s employer, as their parents are more likely to provide access to jobs at high-paying firms. [17/19]

Putting these results together implies that the IGE would be 7.2% lower if no one found jobs through these parental connections. [18/19]

Connections at the parents' employer are not the main driver of intergen mobility. But magnitude (+19%) and source (higher-paying firms) of earnings gains suggest that connections in the labor market (more broadly defined) might an important determinant of mobility. [19/19]

Thanks for reading! You can check out the full paper here: matthewstaiger.github.io/matthewstaiger…