Why the Raw Wage Gap Flips
Young firms often look like lower-paying employers when their workers’ wages are averaged without adjustment. This chapter diagnoses why that symptom appears, separating the firm’s age from the size, sector, worker composition, and tenure patterns that accompany it. The practical conclusion is careful but important: once comparable workers and employment histories are brought into the comparison, the evidence from France and Portugal shows no economically meaningful young-firm pay penalty, and sometimes a small premium—not proof that every young firm pays more.
The symptom: young firms pay less on average
The raw numbers are difficult to ignore. In the OECD data, the average gross hourly wage is about €14.04 in young French firms and €16.26 in older firms. In Portugal, the corresponding figures are €4.51 and €5.38. A simple regression with firm age and year effects produces a wage gap of roughly 10% in France and 12% in Portugal. If the question is only which group of observed workers receives the higher average wage, older firms win.
That simple regression has a legitimate descriptive use: it records the difference a reader would see without asking why it exists. Its coefficient is not yet the effect of changing an employer from old to young. There is no experiment that randomly assigns workers to firms of different ages, and the age groups are built from different firm populations. The coefficient mixes the age of the firm with the characteristics that tend to accompany entry, survival, small scale, and early-career employment. Diagnosis begins by preserving the symptom while refusing to mistake it for the cause.
But that is not yet the question raised by the young-firm mechanism. The raw comparison combines the employer with the people who work there and with the stage of their employment. A young firm may pay less because it employs more entrants, not because it offers less for the same worker. It may pay less because it operates in a lower-paying sector or is much smaller. It may pay less because its workers have had less time to accumulate tenure, even if the wage schedule for a worker with comparable experience is no lower.
The composition differences are large enough to make this more than a technical footnote. In France, firms five years old or less account for about 26% of firms but about 16% of workers; their average size is roughly 14 employees, compared with 44 at older firms. In Portugal, young firms account for about 27% of firms but only 12% of workers, with an average size of about six employees versus 13 at older firms. The workers in young firms are also younger: the French average is about 35.6 years compared with 37.1 in older firms, and the Portuguese average is about 37.5 compared with 40.1.
Sector mix adds another layer. In France, around 87% of workers at young firms are in non-financial market services, compared with about 80% at older firms, while manufacturing has higher average hourly wages than services. Portugal has a similar concentration of young firms in services, although the wage ordering across manufacturing and services differs from France. An age coefficient that ignores sector can therefore absorb the economic structure in which young firms are more likely to operate.
Other job attributes can move the average in the same direction. Young French firms have more part-time workers and a larger share of workers on fixed-term contracts in the descriptive data; Portuguese young firms show especially large differences in fixed-term contracts. These attributes are not interchangeable with wages, and they do not by themselves show that the employer offers a worse job. They do show why the observed workforce must be described before an age comparison is interpreted.