skip tocontent

ombudsman news

issue 37

May/June 2004

insurance - keys left in cars: a continuing problem

Almost all motor policies include a clause that excludes cover for theft or attempted theft if the ignition keys were left in - or on - the vehicle. As this constitutes a major restriction on the scope of cover, insurers need to draw it to the attention of prospective customers, in accordance with the Association of British Insurers' Code of Practice for General Insurance Business. If an insurer cannot demonstrate that it did this, then we are likely to uphold the complaint.

left unattended-

Since the Court of Appeal's judgment in Hayward v Norwich Union Insurance Ltd, many insurers seem to have reworded their clauses to exclude cover for theft if the vehicle was left unlocked and unattended, or if the keys were left in or on the vehicle. This reduces the scope for disputes of fact as to whether the keys were actually in or on the car: strictly speaking, it is enough that the car was left unlocked and unattended for cover to be excluded.

The practical result is that, in many cases, we simply have to decide whether the unlocked vehicle (with or without its ignition keys) was "left unattended".

The leading case on unattended property is still Starfire Diamond Rings Ltd v Angel (reported in 1962 in Volume 2 of the Lloyd's Law Reports at page 217). In this case, Lord Denning (in the Court of Appeal) held that - for a vehicle to be "attended" - "there must be someone able to keep it under observation, that is, in a position to observe any attempt to interfere with it, and who is so placed as to have a reasonable prospect of preventing any unauthorised interference with it". He emphasised that it is a question of fact in each case as to whether the vehicle has been "left unattended".

We think this test is very similar to the one applied by the court in Hayward v Norwich Union Insurance Ltd, where the keys were "left" if the driver moved so far from them that it was unlikely he or she would be able to prevent the theft. Indeed, Lord Denning and his fellow judges did not state that the property/car had to be constantly in view in order to be "attended". The "Starfire Diamond Rings Ltd v Angel" test is not concerned simply with the policyholder's actual observation of the property. It is a theoretical test to ascertain their physical proximity to the property: was the driver close enough to be able to keep the property/car under observation-

In deciding whether a driver was close enough to the vehicle to make a theft unlikely, the location of the incident is important - arguably more so than the physical distance between the driver and the car. After all, what is reasonable in one's own driveway may be unreasonable in public areas where crime of this sort is prevalent, such as petrol stations, recycling units etc.

Having said that, if a driver is standing right next to their car, their mere presence may have a deterrent effect and make a theft unlikely, even if the driver is not physically able to prevent a theft. Indeed, the Court of Appeal recognised this sort of scenario in Hayward v Norwich Union Insurance Ltd, citing the example of a thief making a move while the driver takes something out of the car boot or attends to a child in the back seat. Insurers have sometimes concluded that the mere fact that a theft has occurred demonstrates that the policyholder was not in a position to intervene, but that is not the legal position. What has to be established is whether the driver was in a position to intervene, not whether they were successful in preventing a theft.

We still expect firms to pay claims where the policyholder has not "left" the car. However, some of the more tightly-worded policies mean it may be more difficult for some policyholders to demonstrate factors such as their proximity to the vehicle, observation of it, prospect of intervening, etc. For example, policyholders who have merely turned their back on the car while closing the garage door are likely to succeed; those who have gone indoors to fetch something are likely to fail.

Because of the endless variety of scenarios that occur, these cases can be challenging, particularly when a car is on, or close to, private land - but has been left unlocked or with the ignition keys in it. Typical examples include the situation where the driver has:

  • returned indoors to fetch something;
  • left the car at the bottom of a drive while delivering a package; or
  • left the engine running in order to defrost and demist the car on a cold morning.

As a general rule of thumb, we take the view that a car was "left" if it was actually on the public highway - however close to the driveway or private property - and the driver (and any other responsible person) turned their back on it and walked away from it.

acting recklessly-

Some policies, particularly older ones, do not contain a "keys in car" exclusion clause. Where this is the case, firms may try to reject claims on the basis that the policyholders were in breach of the policy condition that requires them to take "reasonable care". But in order to establish this, firms need to show that the policyholders were "reckless" - in other words, that they recognised the risk but deliberately "courted" it.

People "court" risk if they either take no measures at all, or take measures that they know will not be adequate to avert the risk. This is the test of "recklessness" as set out in the leading legal case on conditions regarding "reasonable care": Sofi v Prudential Assurance (reported in 1993 in Volume 2 of the Lloyd's Law Reports at page 559).

Most people who leave their keys in the car simply fail to recognise the risk and/or take no precautions whatsoever. It is very difficult in these circumstances for firms to show that the policyholders were reckless. If the policyholders had been aware of the risk, they would probably not have left the keys unattended in the first place.

We do not usually need to apply the Sofi v Prudential Assurance test of recklessness in cases involving a "keys in car" or "left unattended" exclusion clause.

However, the tighter the wording of the exclusion, the more onerous or unusual the exclusion is likely to be - and therefore the greater the insurer's obligation to highlight the precise terms of the policy. We know from experience that consumers are frequently unaware that such an exclusion forms part of the policy terms. If an insurer attaches an unusually restrictive term to a policy, then it must make sure that anyone considering buying such a policy realises that the theft cover is unusually limited. Ideally, we would like to see these sorts of restrictions clearly highlighted on the policy certificate (which customers have to possess by law) and on the policy schedule (which is the document that customers are more likely to read).

In our next issue, we will illustrate how we put these general principles into practice.

Walter Merricks, chief ombudsman

ombudsman news issue 37 [PDF format]

ombudsman news gives general information on the position at the date of publication. It is not a definitive statement of the law, our approach or our procedure.

The illustrative case studies are based broadly on real-life cases, but are not precedents. Individual cases are decided on their own facts.