The precise timing of a family crisis can shape a child’s future. In new research co-authored with Kjell Salvanes, Barton Willage and Alexander Willén, we look at how a parent losing a job affects children at different ages, from toddlers to teenagers (Carneiro et al, 2026).
By studying decades of data from Norway, we identify a striking pattern: parental job loss significantly harms children exposed in early childhood, and harms them even more when it happens during the teenage years – while children exposed in the middle, roughly ages 6 to 10, are barely affected at all. This suggests that adolescence is, alongside early childhood, a particularly sensitive period, one in which family stress can derail a child’s path to success.
Timing is everything: evidence from Norway
We use records from the entire population of Norway between 1986 and 2018 to track how families were affected by economic ‘shocks’ such as mass layoffs and factory closures. These events are useful for research purposes because they are usually outside a parent’s control and happen regardless of how old their children are. This makes it possible to compare children who were similar in every way except for the age that they had reached when their parent lost a job.
The timing of the job loss matters immensely. If a parent is laid off when a child is 11 to 16 years old, the negative consequences are the largest: these teenagers end up with lower grades, are less likely to graduate and go on to earn less money by the time they turn 30. For children aged 0 to 5, there is also a meaningful negative effect, although on most measures it is somewhat smaller than the effect on teenagers. Surprisingly, for children in the middle – ages 6 to 10 – a parent losing a job has almost no effect on their long-term outcomes.
Another key discovery is that it matters which parent loses the job: maternal job loss has a much larger negative effect on children than paternal job loss. And this doesn’t appear to be simply because mothers lose more income – income losses were, if anything, similar or slightly larger for fathers, especially when the children involved were teenagers. This points away from money as being the main story, and towards how job loss changes a mother’s wellbeing and the atmosphere at home.
Parental employment and children’s development
For a long time, many experts believed that the first few years of life were the critical window for children’s development. Early childhood is certainly important, but our study suggests that adolescence is a second – and, on several measures, comparably or even more consequential – window. During the teenage years, children are dealing with rapid brain development, social pressure and high-stakes exams. A major source of stress at home during this time can leave a lasting mark on their trajectory.
The findings also change how we think about money versus atmosphere. Because maternal job loss hurts children more than paternal job loss – despite not costing the family more in income – our research points to the emotional environment of the home as an important driver. When mothers lose their jobs, the data show a rise in their stress, anxiety and sleep problems in the years immediately afterwards. Teenagers exposed to a parent’s job loss show a similar rise in their own diagnosed anxiety, stress and sleep problems, suggesting that the disruption reaches children directly and not only through their parents’ distress.
We also find suggestive evidence that shocks can pile up. For outcomes like grades and the quality of the school programmes to which children are admitted, a child who experiences a job loss at home as a toddler and then another as a teenager appears to fare worse than you’d expect from simply adding up the two shocks. That said, this compounding pattern isn’t universal: it shows up much less clearly for outcomes like whether a child graduates or enrols in college. But where the pattern does appear, it suggests that children who are already carrying one disruption are especially vulnerable to a second one.
What this could mean for policy
Our study wasn’t designed to test specific policy interventions, and the paper itself is appropriately cautious on this front – it notes only that the timing of a shock, and not just whether it happens, matters for how social insurance and family policy might be designed.
Still, a few implications seem worth exploring further. Unemployment benefits and family support programmes today are typically the same regardless of a child’s age. Our results suggest that it may be worth studying whether extra support for families with teenagers or very young children – the ages where the damage looks most likely to stick – would make a measurable difference.
It’s also worth asking whether cash support alone is enough. Because the emotional environment of the home, rather than the household budget, seems to be central to how job loss affects children, there may be a case for schools and healthcare providers paying closer attention to family mental health when a major local employer closes down – for example, by helping mothers to manage stress and sleep after being laid off. This is a hypothesis towards which our results point, not something that we test directly, so it deserves its own evaluation before being built into policy.
Finally, our results suggest one candidate explanation for why teenagers are so vulnerable: adolescence coincides with a run of high-stakes decisions in the Norwegian school system – the grades that determine which upper-secondary track a child can enter, and from there, access to university. A disruption may do lasting damage precisely because it lands during this narrow, consequential window, in a way that it might not if it happened a few years earlier or later. Testing this explanation directly, and understanding whether it holds in school systems that are organised differently, is a natural next step for research.
