There's Life in the Old Dog yet … and in Car Insurance!
Car Insurance in the course of time
For some time now, leading voices in the insurance industry have been singing a swan song about car insurance. Consultants, consumer protection specialists, politicians, market analysts and even the insurers themselves are joining in. The chorus goes like this: Driver assistance systems will automate driving and make the vehicle «autonomous». When the human factor is removed, there will be no more accidents and the end of car insurance will have arrived. There may be some truth in this, but the claim ignores certain realities. For one thing, car insurance is also being driven by technological change and other ongoing developments. In conclusion, there have always been technological adaptations – and that is anything but being weary of life!
Fears versus facts
Anyone who has dealt with car insurance in recent years is aware of the claims and studies which predict the slow death of this business. That would be a catastrophe for established insurers, as the revenue from motor liability and comprehensive insurance accounts for almost 40% of total non-life business revenue.
The reason for the slow decline, the doomsayers say, is technological progress in automobile manufacturing and the resulting autonomy of vehicles. Today, human error is behind 90% of all accidents. But in a world of completely autonomous driving, accidents would no longer occur, and car insurance would no longer be needed, these doomsayers say. This sounds plausible, but it is wrong. The logic is not compelling, and the underlying predictions are pure speculation.
Let's look at the facts: Vehicles are increasingly being equipped with driver assistance systems. In addition to parking assistants, there are emergency braking systems, lane change warning systems, as well as pilot systems for motorway driving. Manufacturers are working hard to further develop these systems. There are already breathtaking examples of what is technologically possible today, and they give us an idea of what the future of the automobile will look like. However, we tend to overestimate the effects of new technologies. This is because we like to extrapolate developments or tendencies linearly, so a shift to autonomous driving seems logical.
The notion of the self-driving car has led to a hype that ignores three essential factors: The time of development, the legal framework and international traffic. We take a look at these factors below.
The factor time
The first thing to bear in mind is what time periods we are talking about. Let us take a look at the largest market in Europe: Germany.
For decades we have had a continuously growing vehicle population there. It currently totals around 50 million passenger cars. The «life expectancy» of a car in Germany is around twelve years. The average age of a car there is almost ten years and is rising. This means that vehicles registered today will be on the roads for more than ten years.
The number of assistance systems being equipped on cars from the luxury class down to the medium and small car classes is increasing. There are also voluntary retrofits for older vehicles. Compulsory retrofits such as those for safety belts or the compulsory decommissioning of unassisted vehicles are unlikely. It is therefore easy to calculate how long it will take before we have complete penetration of assistance systems in cars. In our view, a completely autonomous vehicle population will only become reality after 2040
The legal framework
The second factor is that the owner of the vehicle is obligated by law to take out and maintain liability insurance for the vehicle. It would be naïve to expect that to change in the medium term, but that is what the aforementioned studies have done. They argue that if the person no longer drives, the liability shifts from the owner to the product provider (i.e. the vehicle manufacturer or the provider of the assistance system). This sounds plausible but is actually wrong in this context. Product liability exists of course but is almost irrelevant in cases of vehicle accidents.
The German insurance industry association (GDV) does not expect widespread introduction of autonomous driving before 2035. This date can be regarded as very optimistic for the aforementioned reasons. The GDV also expects piecemeal introduction of self-driving technology, for example only on certain motorway sections. This implies that we will have a longer transitional period of mixed traffic, that is vehicles that are assisted and those that are not. Mixed systems are more susceptible to disruption than closed systems consisting either of equally assisted vehicles or those including pedestrians, cyclists and motorcyclists.
Experience shows that revolutionary technologies lead to discussions in our society. That tends to delay innovation. In Europe, innovations are not decided by decree. This will likely make autonomous driving possible in Asia before it comes to Europe.
International cross-border traffic
Finally, we often forget that our transport routes are international networks. It thus makes little sense to promote national transport solutions. We must also assume that there will be a longer transitional phase with mixed traffic, because, as we have seen, pan-European solutions take a long time. We have discussed some legal and technological aspects of autonomous driving that are in the public eye. We will now discuss the implications for underwriting.
The underwriting perspective
In 2016, the GDV and a research institute tied to the Allianz Group published a study showing that assistance systems led to a decline in the number of accidents. But the decline was significantly less than one would expect. Moreover, a negative consequence emerged, namely that the damage can become more expensive. If, for example, sensors and cameras must be replaced along with broken glass, the entire system must be recalibrated. The additional cost varies from manufacturer to manufacturer, but the GDV puts it at an average of 30%.
The study concludes that assistance systems could reduce car insurance compensation by between seven and 15%, depending on the speed of distribution. The declines differ for liability and hull insurance. However, the new technology also creates new risks. These include unforeseeable system failures and possible hacker attacks. These should not be underestimated, as they would not affect individual vehicles, but rather several of them or even an entire model series. That would lead to considerable damage costs that are generally not covered by today's policies.
That said, car insurers are aware of these new risks and the challenge of accounting for them. We still have a long way to go before autonomous driving is widespread. On the way there, car insurance will change. But it will certainly survive.