Development compared to other age groups
Figure 9: Driver fatalities per million population for different age groups in the period 1990-2003 in OECD countries
In comparison with other age groups, the fatalities for young drivers have decreased to about the same extent as for the age group 25-54 (Figure 9).
Development per country between 1985-2003
Has road safety of young drivers improved over time? To answer this question Figure 10 compares the mortality rate of young drivers in several countries in two periods.
Figure 10: The development of driver fatalities per million population of age group 18-24 comparing two periods: 2000 & 2003 and 1985 & 1990 Source OECD, IRTAD
Overall, a clear pattern shows that in most countries the fatality rate has gone down over time. Some countries are an exception to this general rule. It seems to be those countries that have experienced sharp increases in motorization in a relatively short period of time, like the Czech Republic, Poland and Ireland. But Norway, the Netherland and the UK all show very little or no change. However, a detailed comparative analysis of Sweden, The Netherlands and the United Kingdom  also shows the limitation of the analyses of fatalities per head of population presented here. When exposure is taken into account and fatalities are calculated per distance driven, these three countries also show an improvement over time.
Young driver risk and general safety levels
It is a relevant question to what extent young driver risk is related to general safety levels in a country. Do safe countries have relatively safe young drivers? And the other way around: Do unsafe countries have relatively unsafe young drivers?
This question is relevant to understand the nature of the problem. If, all young drivers have problems irrespective of the safety level of a country, then specific measures targeted at young drivers are the only way forward. Similarly, if general safety level is dominant factor, then improving the general safety levels is the preferred action.
Figure 11: Fatalities per head of population per country in two age-groups compared by the OECD average
In Figure 11 the risks of experts and novice drivers are compared between countries. Each country has received two scores. One score on the novice driver risk and one for the expert driver risk. The white lines indicate the average value for the two groups in the OECD: the white horizontal line for the expert drivers; the white vertical line for the novice drivers.
The figure shows that, in general, countries with safe expert drivers have relatively safe young drivers as well (quadrant low old/ low young). The opposite is also true (quadrant high old/high young). This indicates that, the improvement of general safety levels (e.g. by enforcement, infrastructure, injury protection) is a valuable way toward improving the safety of young drivers.
But above all, this graph also indicates that in most countries, young drivers have elevated risk levels. This indicates that in addition to the general safety measures there is a strong need for specific young driver measures as well. Judging by the absolute differences between general safety levels within counties for young and older drivers, it may be worthwhile to begin with those measures that improve the situation for all drivers, and to focus on the individual risk groups such as young drivers after that.
The special case of young male drivers
Some recent findings indicate that the young male driver increasingly becomes a problem and does not respond to measures to the same extent as female young drivers do.
Figure 12: Relative risk ratios of young male young female drivers over the past 10 years.
The reference group is the group of older more experienced drivers of both sexes (30-59)
In a comparative study between the United Kingdom, Sweden and the Netherlands  the risk for young drivers of a serious accident per kilometre driven was calculated for each year in the period 1994-2003. This risk was then compared to the risk of expert drivers in those countries, resulting in a relative risk ratio. Values higher than 1 indicated a worse safety record for novice drivers. Figure 12 shows that female drivers have a worse record than expert female drivers, but that in all three countries the ratio is roughly constant over this time period. The relative risk ratio for young males was already higher than for females in the mid 1990s, and does not respond to overall improvements in safety levels. This pattern is consistent in all three countries. Therefore, we conclude that special attention needs to be paid to this trend.