Table 3 contains information on the wage share of workers under age 18 for various definitions of low-skilled work. About 2.8% of employed individuals are under age 18, and due to the fact that younger workers work fewer hours on average than older workers, the share of employment in hours is even smaller: 0.8%. Focusing on wage shares to match our theory, we find that the overall wage share is 0.3 for young workers. The last column of table 3 reports the lower bound for the elasticity of youth employment with respect to the youth minimum wage obtained from plugging in the given wage share as the value of $δ$ and our baseline estimate for the wage elasticity from the discontinuity ($ε=0.82$). Using a wage share of 0.3% results in no meaningful change in the elasticity. However, the theory that gives the bounding result suggests that we should be using the wage share for low-skilled work that is perfectly substitutable for work by employees under age 18. Various approaches to pin down this number lead to a higher wage share, as shown in table 3. One approach is to use a concept of low-skilled workers that includes workers similar to young workers, based on their sector (e.g., supermarkets), their hourly wage rate, or their education. These definitions lead to wage shares ranging from 1% to 7%, which are once again small enough that they have little effect on the implied elasticity. A more conservative method for defining low-skilled work is to suppose that only younger workers are perfectly substitutable for workers under age 18. In an extremely conservative calculation, in which we assume the only substitutes for 16- and 17-year-olds are 18- and 19-year-olds, we obtain a wage share of 26.5%. This conservative wage share leads to a lower bound of the true elasticity of youth employment with respect to the youth minimum wage of 0.60. With this lower-bound elasticity, increasing the youth minimum wage up to the level of adults would decrease youth employment by 24% instead of 33%, still a sizable employment effect.

Table 3.
The Share of Younger Workers in the Low-Skilled Labor Market
PopulationShare Age 16–17 (%)Lower-Bound Elasticity
Full population 4.0
Employment (persons) 2.8
Employment (hours) 0.8
Wage income 0.3 0.82
Low-skilled occupations$a$ 2.1 0.80
Supermarkets 6.5 0.77
Hourly wage less than 95th percentile for 18-year-olds$b$ 1.0 0.81
Highest education nineth grade or lower$c$ 2.2 0.80
Individuals age 16–24 5.6 0.77
Individuals age 16–19 26.5 0.60
PopulationShare Age 16–17 (%)Lower-Bound Elasticity
Full population 4.0
Employment (persons) 2.8
Employment (hours) 0.8
Wage income 0.3 0.82
Low-skilled occupations$a$ 2.1 0.80
Supermarkets 6.5 0.77
Hourly wage less than 95th percentile for 18-year-olds$b$ 1.0 0.81
Highest education nineth grade or lower$c$ 2.2 0.80
Individuals age 16–24 5.6 0.77
Individuals age 16–19 26.5 0.60

This table reports the wage share of workers aged 16 and 17 in selected populations, providing suggestive evidence on the share of younger workers in the low-skilled labor market, and computes the corresponding lower-bound elasticity estimate using formula (9). In the baseline calculations, we use data for all Danish employees ages 16 to 65. In the last two rows, we assume that only workers of ages 18 to 24 and 18 to 19, respectively, may substitute for young workers.

$a$We identify low-skilled occupations using the four-digit ISCO classification. We select the ten most important occupations or job types for youth, which correspond to about 83% of youth employment.

bWe define low-skilled adult workers as having a wage below the 95th percentile for 18-year-olds.

$c$We count low-skilled workers as all workers over age 18 with an education level of nineth grade or lower, together with all 16- and 17-year-old workers.

Close Modal