Young men, risk, and the worry gap

In 2016, Pierluigi Cordellieri and colleagues at Sapienza University of Rome published a study in Frontiers in Psychology that did something unusual for this field: instead of surveying one country, it pooled 2,681 young road users aged 18 to 22 from nine European countries, Italy, Austria, Bulgaria, Cyprus, Germany, Ireland, Latvia, Lithuania and Poland. The sample was 1,458 male (54.4%) and 1,223 female. Each person filled out standard scales: the Driving Attitudes Scale for attitudes toward rules, speeding and drink or drug driving, the Driver Behaviour Questionnaire for self-reported violations and errors, and two single questions on accident risk, one asking them to rate their own probability of a crash compared with people their own age, the other asking how worried they were about that possibility.

Now the honest qualifiers, because they change how you should read everything that follows. First, these are described as young road users, not all drivers. Participants were sorted into three groups, car drivers, motorcyclists and non-drivers, and the behaviour scales were adapted so that non-drivers answered about imagined or intended driving rather than logged experience. So the behaviour findings are partly about what young people say they would do, not only what they have done. Second, the two most important measures, risk and worry, are single items each on a ten-point scale, not multi-question instruments, which makes them blunter than the attitude scales. Third, Frontiers in Psychology is a respectable but mid-tier journal, and the study is cross-sectional self-report, so it can show associations, not causes. None of this sinks the work. It just means the headline is a signpost, not a verdict.

One more piece of context from the paper's own introduction, because it guards against a misreading we are about to warn you off. The authors are explicit that young drivers as a group tend to underestimate their crash risk and overestimate their skill. That is the backdrop. So nothing here says young people, male or female, judge road risk perfectly. The finding is narrower and more interesting than that, and getting the wording exactly right is what the next two sections are for.

The researchers put the two risk items into a single analysis crossing gender with measure. Two results came out of it. On the first item, the rating of your own crash probability compared with people your age, young men and young women did not differ significantly. On the second item, worry about that same possibility, they differed sharply: young men were significantly less worried, F(1,2652)=115.552, p<0.001, with a moderate effect size, Cohen's d=0.42. The interaction that formalises the split, gender behaving differently across the two measures, was itself strong, F(1,2650)=101.546, p<0.001. In plain terms: asked to judge the odds, the sexes answered alike; asked how much it bothered them, the young men shrugged where the young women did not.

Why does a gap in worry, rather than in judgement, matter at all? Because worry and risk perception are not the same psychological thing, and the paper leans on two of its cited sources to say so. Sjöberg (1998), as cited by Cordellieri and colleagues, frames the distinction as emotion versus cognition: perceived risk is a cognitive assessment of vulnerability, while worry is the affective response to a threat. And Peters and colleagues (2006), also cited in the paper, reported that worry predicted precautionary intentions better than risk perception did. If that holds, then two people can hold the same cold estimate of danger and behave very differently, because one of them feels enough about it to act and the other does not. That is the mechanism the authors reach for. It is worth being clear that they reach for it as an idea, not a proof, which is exactly what the cross-examination takes up next.

Alongside the risk items, the study found a consistent set of attitude and behaviour gaps, all significant across the pooled sample. A note on how we report them: the paper prints Cohen's d values, but the signs of those d values are internally inconsistent in the tables, so we describe each gap by its magnitude and direction in words and give the d only as a size. Two of the country-level tables also contain obvious typographical errors, values like d = -5.94 that cannot be real effect sizes, so we do not quote any per-country d at all and describe the cross-country pattern qualitatively instead.

Two things in that table deserve a second look. First, the drink-and-drug result runs the other way to the rest: young women were the group more firmly opposed to driving impaired, which lines up with the paper's cited evidence that women are less involved in alcohol-related crashes. Second, and more interesting, young men reported more inattention errors as well as more violations. The neat old textbook division, in which men deliberately break rules while women make honest mistakes, does not survive this data. Here the young men were higher on both. That matters because it means the young-male pattern is not simply "rule-breaking by choice"; some of it looks like less careful driving overall, which is not something a poster about consequences addresses.

On geography, the paper is careful, and so should we be. The anti-rule attitude gap was significant in all eight analysable countries, Ireland excluded for having no men in its sample, which makes it the study's most strongly supported cross-cultural finding. The speeding-tolerance and drink-driving gaps, by contrast, were not significant everywhere. The authors read that as culture and policy showing through: how much young men out-tolerate young women on speeding or drink driving depends on the country's drinking norms and enforcement, not just on being young and male. So even the sturdiest part of this study comes with a reminder that context does a lot of the work.

Start with the distinction that keeps this honest. The kind of concern worth building is internally owned and forward-looking: a young driver who has genuinely thought through what could go wrong on this road, at this speed, with these passengers, and acts on it before anything happens. That is worlds away from a shock imposed from outside in a thirty-second ad. Sweeny (2017), reviewing what she calls functional worry, argues that a moderate, action-oriented concern is protective: it cues that a situation is serious, keeps the risk in mind, and motivates people to reduce it, and it predicts more health-protective behaviour such as cancer screening. She also describes an inverted-U: too little worry leaves the over-confident young driver taking no precaution, but too much tips into paralysis and avoidance. The goal is never to maximise negative emotion. It is to move someone from too little concern to about the right amount, and no further.

For a parent or an instructor, the translation is practical. The worry gap is not closed by a lecture about death, which this group is unusually good at deflecting. It is closed by putting a young driver in situations where the risk becomes real and owned: unhurried commentary drives where they say aloud what could go wrong and when they would act, honest debriefs after a near-miss instead of a telling-off, structured practice against the learner logbook, and calm exposure to the conditions, wet nights, tired evenings, a car full of mates, that the statistics single out. That is engagement, and it is what our teen and novice drivers guide covers for the developing brain and the parent's role, and what our careless or clueless review covers for the underlying mechanism. If you are a parent weighing up how to set your young driver up well, our page for parents and sponsors is the next step.

The strongest clean Irish figure comes from the RSA via a parliamentary question, reported in May 2026: from the start of 2024 to 20 May 2026 there were 101 road deaths in collisions involving a male driver aged 16 to 25, 39 in 2024, 47 in 2025 and 15 in 2026 to late May. In 2025 those deaths made up 26% of the 179 fatalities recorded by Gardaí that year. Read the metric carefully: this counts deaths in collisions involving a young male driver, victims of any age, not deaths of young men. Of the 2025 young-male-driver-involved deaths, 22 were the drivers themselves, 13 passengers, 8 pedestrians, 1 motorcyclist and 3 cyclists. The RSA's framing is qualitative but consistent: of roughly 190 deaths in 2025, about 74% were male, the 16-to-25 band is the highest-risk group, and young men in their twenties are over-represented, typically on rural roads, which the RSA attributes to overconfidence and a sense of invincibility.

One place the worry gap connects to Irish rules cleanly is the novice period. A driver is a novice for the first two years of holding a full licence, must display red N plates front and rear for those two years, and is held to a tighter penalty-point threshold: seven points in three years brings an automatic six-month disqualification for learner and novice drivers, against twelve for experienced ones, and the drink-drive limit is the stricter 20mg per 100ml. Those first months after the test are statistically the most dangerous stretch a driver ever has, which is exactly when engagement matters most and a false sense of having "passed, so I'm good now" is most costly. Our guide to N-plate and novice rules sets out the full detail, and any precise fine should be checked against the live RSA page, since legal amounts change.