How many lives could be saved if everyone complied with the speed limit? – A case study from Sweden

Increased speed compliance and reduced speeds are important factors in achieving increased traffic safety since it affects both the number of crashes and the injury severity. Lower speeds reduce the reaction distance leading to greater opportunities to avoid a crash, to reduced impact forces and to less severe injuries in the crash. The aim of this study is to investigate how many lives that could be saved if all drivers complied with the speed limit on the Swedish road network. The study is based on results from extensive speed measurements done on the rural and urban road network in Sweden and crash statistics from the crash data base Strada (Swedish Traffic Data Acquisition). To estimate the traffic safety effects the Exponential model is used. The results show that if everyone kept the speed limit in Sweden, about 50 lives could be saved yearly representing 25 % of all fatalities.


Background
Speed and its management are the core in a Safe System approach and cuts across most Safe System intervention categories.In a Safe System, three main strategies to eliminate fatalities and severe injuries due to road crashes can be identified: prevent people from exposure to risks, reduce the risk level, and protect people from harmful impact forces in the event of a collision.Managing and controlling speed is therefore at the center of developing a Safe System.The objective of managing speed in a safe system includes, besides protecting road users from severe injuries and deaths, a desire to create a sustainable society addressing livable cities, health, security, and equity.____________

Background
Speed and its management are the core in a Safe System approach and cuts across most Safe System intervention categories.In a Safe System, three main strategies to eliminate fatalities and severe injuries due to road crashes can be identified: prevent people from exposure to risks, reduce the risk level, and protect people from harmful impact forces in the event of a collision.Managing and controlling speed is therefore at the center of developing a Safe System.The objective of managing speed in a safe system includes, besides protecting road users from severe injuries and deaths, a desire to create a sustainable society addressing livable cities, health, security, and equity.____________ Safe speeds and safe speed limits are on the global agenda.In advance of the Third Ministerial Conference on Global Road Safety 2020, an Academic Expert Group, consisting of road safety experts from around the world, developed nine recommendations to advise on priority directions for road safety following the first Decade of Action, Tingvall et al. (2020).These recommendations are based on the introduction of Agenda 2030, often referred to as the Sustainable Development Goals (SDGs).One of the recommendations is about speed compliance, namely Recommendation #7: Zero Speeding.Reduced speeds and increased speed compliance are important factors in achieving increased traffic safety.Reducing the speed of the traffic, affects both the number of crashes and the injury severity in several ways.First, lower speeds reduce the reaction distance so that you have greater opportunities to avoid a collision or crash, second, lower speeds reduce the impact force in the crash, which makes the injuries less severe (ITF, 2018).Even small changes in driving speeds could have a substantial effect on road safety.As a rough estimate, a 10% decrease in mean speed leads to 20% decrease in injury crashes and a 40% decrease in fatal crashes (ITF, 2018based on Elvik, 2013).
In Sweden there is a national goal that the number of fatalities should be reduced by 50% and the number of severely injured by 25% during the period 2020 -2030.Speed compliance is low, especially on the rural network where only about 45% of drivers keep the speed limit (Greijer och Nyfjall, 2020).Increased speed compliance has been identified as one of the most important areas to achieve these traffic safety goals.
The aim of this study is to investigate how many lives that could be saved if all drivers complied with the speed limit on the Swedish road network.

Method
To estimate how many lives that could be saved if everyone kept the speed limit in Sweden, several assumptions are made: • The speeds per speed limit follow a normal distribution.
• The mean and standard deviation in the normal distribution is estimated based on results from national speed surveys.• A change in speed compliance to achieve zero speeding is assumed to happen in the following way: Everyone exceeding the speed limit reduces their speed to exactly the speed limit, drivers keeping the speed limits do not change their speed.
In addition to these assumptions results from extensive speed measurements done on the rural and urban road network in Sweden are used, as well as crash statistics from the crash data base Strada (Swedish Traffic Data Acquisition).These data are described below.

Speed data
Space mean speeds and standard deviations (SD) used in the analysis are shown in Table 1.These mean speed levels are based on speed surveys done on the Swedish road network in 2020 (Greijer och Nyfjall, 2020 andVadeby andAnund, 2021) and describes the speeds per speed limit in the before situation.
Table 1.Space mean speed and standard deviation (SD) based on measurements done 2020.

Data on fatalities
Data on fatalities are official statistics from the crash data base Strada (Swedish Traffic Data Acquisition) and in the analysis, a three-year average from 2018 -2020 is used.The number of fatalities per speed limit is shown in Table 2.For about 5% of the fatalities, the speed limit at the location of the crash was not known.For these crashes, it was assumed that these fatalities follow the same distribution over the speed limits as the fatalities with known speed limits.The mean number of fatalities during 2018 -2020 was 249 per year.
Table 2. Mean number of fatalities during 2018 -2020 per speed limit.All speed limits 249

The Exponential model
A change in speed compliance leads to a change in mean speed.To estimate the traffic safety effect due to a change in mean speeds on the entire road network in Sweden, the Exponential model developed by Elvik et al. (2019) is used.The Exponential model describes the relationship between a change in mean speed and the related change in fatalities or injured by means of an exponential function: (1) where vbefore and vafter is the mean speed before and after the change in speed compliance.The coefficient β = 0.08 (SD = 0.003), see Elvik et al. (2019) and ybefore and yafter are the number of fatalities before and after the change of speed compliance.

Analysis
To achieve that everyone keeps the speed limit, a change in speed compliance is assumed to happen in the following way: Everyone exceeding the speed limit reduces their speed to exactly the speed limit, drivers keeping the speed limits do not change their speed.To estimate possible effects, the following analyses are done: 1. Simulate normal distributions with mean speeds and standard deviations according to Table 1.N = 1 000 000 vehicles for every speed limit.This gives the speed distribution in the before situation.
2. For all speed limits, change the speed distribution by changing all speeds above speed limit to the speed limit.All other speeds remain unchanged.This gives the speed distribution in the after situation.
3. Calculate space mean speed based on the changes described in 2 above.This gives the mean speed in the situation that everyone keeps the speed limit, i.e. vafter.
4. Calculate the change in mean speed, (vaftervbefore) 5. Calculate estimated change in the number of fatalities by means of the Exponential model.This is done for every speed limit and then summarized to get at total number.

Results
By simulation of speeds from the normal distributions (N = 1 000 000 for every speed limit) and assumptions that everyone keeps the speed limit a change in the speed distribution is achieved.An example of the change in the speed distribution for speed limit 80 km/h is shown in Figure 1.The mean speed in this example is changed from vbefore= 78.9 km/h in the original before situation to v after = 74.9km/h.Figure 1.Example of a speed distribution.The speeds follow a normal distribution at speed limit 80 km/h with mean speed vbefore= 78.9 km/h.Figure 2. Example of a speed distribution, speed limit 80 km/h.Red curve shows current speed distribution with mean speed vbefore= 78.9 km/h and the blue curve shows the speed distribution if everyone kept the speed limit with mean speed vafter = 74.9km/h.
Space mean speed before (based on measurements in speed surveys), space mean speed after (based on simulations and assuming everyone keeps the speed limit) and estimated yearly saved number of fatalities per speed limit are shown in Table 3.Using a three-year average (2018-2020) for the number of fatalities and applying the Exponential model to the speed changes achieved it is shown that 51 lives could be saved yearly.The largest reductions are estimated to appear on roads with speed limit 70, 80 or 90 km/h.Table 3. Space mean speed before (based on measurements in speed surveys), space mean speed after (assuming everyone keeps the speed limit) and estimated yearly saved number of fatalities per speed limit.

Discussion
The aim of this study is to investigate how many lives that could be saved if all drivers complied with the speed limit on the Swedish road network, i.e zero speeding.The results show that based on the level of fatalities during 2018 -2020 in Sweden, about 50 lives could be saved yearly, representing about 25% of all fatalities on the Swedish road network.The largest reductions seem to appear on roads with speed limit 70, 80 or 90 km/h.These are mainly rural 2-lane roads where seldom larger road safety investments are done.ETSC (2019) showed that a high percentage of vehicles in Europe in general exceeded the speed limits on all types of roads.On urban roads, for European countries, between 35% -75% of drivers exceeded the speed limit, on rural roads between 9% -63% and on motorways 23% -59%.Overall, ETSC estimated that about 2100 lives could be saved each year if the average speed dropped by just one km/h on all roads across the EU.
The traffic safety gains of zero speeding are contained from the fact that increased speed compliance leads to a decrease in mean speed and that even small decreases in mean speed leads to decreases in the number of fatalities and seriously injured (ITF, 2018based on Elvik, 2013).When focusing on speed compliance, a very central factor is the speed limits and how they are set in a country.In a safe system, three main strategies to eliminate fatalities and severe injuries due to road crashes can be identified: prevent people from exposure to the risks, reduce the risk level, and protect people from harmful energy in the event of a collision.Managing and controlling speed is therefore at the center of developing a Safe System.Setting speed limits according to safe system principles means that road users' vulnerability and tolerance against external forces should be the main design parameter for the speed.This requires that the speed limits are related to the infrastructure, human vulnerability, and best practices in vehicle design.The safer the roads and vehicles, the higher the speed can be accepted.Setting speed limits according to safe principles means that for urban roads with possible conflicts between cyclists/pedestrians (VRU) and motorised vehicles a default speed limit of 30 km/h is mainly used.While for urban/suburban roads with fully segregated protection for VRU but with remaining intersections risking side impact for cars a speed limit of maximum 50 km/h is used.For rural roads without median barrier protection, risking possible frontal head on collisions, the speed limit should be 70/80 km/h.Only where the roads have median barrier, forgivable roadsides, separation of vulnerable road users and no dangerous intersections, operating speed of 100 km/h or even higher can be considered.
Sweden is a country going in the direction towards speed limits in line with safe system design principles.This change of direction for setting speed limits started in 2004 with a government commission whereby the Swedish Road Administration (SRA) were to present a strategy for gradual adjustment of the speed limits in line with the concept of Vision Zero but also consider accessibility requirements, good environment, regional development, and a gender equal transport system.This has led to major changes of the speed limits on the Swedish roads, Vadeby andForsman (2014, 2018) and many rural roads have got a decreased speed limit from 90 km/h to 80 km/h.The process of adapting the speed limits in line with the concept of Vision Zero is still an ongoing process in Sweden.A decrease in speed limit, often leads to a decrease in speed compliance since the drivers do not fully adopt to the new speed limit.This has been seen in several studies (Vadeby and Forsman 2014, 2018and Cerema 2020), but since the mean speed decreases, the safety effect has been positive.Adopting the speed limits to safe system principles often means decreasing speed limits on some roads where the speed compliance might be worse and thus an even more important issue to consider.In the nine recommendations (Tingvall et al. 2020), recommendation #7 is about zero speeding.To achieve zero speeding, examples of effective measures are: effective enforcement methods and practices, along with substantial penalties for offenders, use of vehicle technologies that detect speed limits and prevent higher speed and use of connected vehicle technology in conjunction with speed limiters and geo-fencing to control speeds in specific areas.In addition, it is suggested that businesses, governments, and other fleet owners practice a zero-tolerance approach to speeding in their own or procured transport operations, that roads are designed so that drivers find it most comfortable to travel at safe speeds and that public education focus on the risks associated with speeding along with awareness of active enforcement activity.
The present study is based on several assumptions and has some limitations.Like a previous study by Sorensen et al. (2007), the speeds are assumed to be normally distributed.Although speeds for a given speed limit do not follow an exact normal distribution, this assumption is considered reasonable in the context.Regarding the model used to estimate the effects on fatalities, the present study is based on Elvik's recommendations regarding the use of the Exponential Model instead of the Power Model and the most recently available exponents (Elvik et al., 2019).
There are also some limitations with data that can be mentioned.There are also no national measurements on streets with a speed limit of 30 km/h or below available.On the state road network, roads with a speed limit of 40 or lower are reported together and on the municipal road network, 30 km/h is not measured in the national survey.Therefore, both fatalities and travel speeds are only studied for the aggregate for speed limits of 40 km/h or below.
The assumption made regarding that everyone driving over the speed limit reduces their speed to exactly the speed limit, while drivers keeping the speed limits do not change their speed creates an asymmetrical distribution with significantly less standard than in the before-situation.It is difficult to assess how realistic this assumption is and how the speed distribution would become in reality.In future studies with e.g., geofencing where the maximum speed is regulated to the speed limit, there might be opportunities to investigate this in more detail.
In conclusion, it can be stated that the study has certain limitations where the most important ones are discussed above but gives a reasonable estimate of the magnitude of the reduced number of fatalities that can be achieved if everyone kept the speed limit.

Conclusion
Increased speed compliance and reduced speeds are important factors in achieving increased traffic safety since it affects both the number of crashes and the injury severity.Lower speeds reduce the reaction distance leading to greater opportunities to avoid a crash and to reduced impact forces in the crash meaning less severe injuries.If everyone kept the speed limits on the road network in Sweden.about 50 lives could be saved yearly, this represents about 25% of all fatalities.This estimation is based on several assumptions and the results should therefore only be seen as an estimate of the magnitude of the reduced number of fatalities.