Devon CC v TR

TR was the driver of a Land Rover on a country road. Whilst he was in the act of overtaking a slower-moving vehicle, his Land Rover left the road to the nearside and his passengers were very seriously injured in the ensuing crash into trees alongside the road. They sued him and he made a third-party claim against the highway authority (“Devon”), alleging that the defective state of the offside of the road was what had caused him to lose control. The claims of the passengers against the driver were settled, leaving the third party claim to be resolved at trial. That claim came on for hearing before Slade J. By her judgment given on 29 March 2012, [2012] EWHC 796 (QB), she found that the highway authority was liable and she held that TR had not been negligent at all. This is the appeal of the highway authority from those two decisions.

The road in question was part of what is mostly a winding and hilly country C-road leading from the A30 Honiton by-pass up into the Blackdown Hills. TR was on his way back from Exeter airport where he had collected one of his passengers. The accident happened about four miles or so after leaving the main road. TR was driving north. He had been following a slower-moving red Vauxhall Vectra. After a winding and steep section of road, and a right hand bend, the road ahead opened up into a straight section which gave ample visibility to overtake. The road was a single carriageway in each direction. At some points it had no line markings at all, but on the section where the accident occurred there was a long dash centre white line and solid white lines down each edge. On both sides there were soft grass verges. On TR’s nearside (west) there was a ditch with trees beyond it, approximately level with the road surface. On his offside (east), there was a noticeable and quite steep short bank down into a ditch, and a hedge beyond it. The weather was fine but it had been raining overnight and there was both some moisture on the surface of the road and puddles in places. It was a midweek early afternoon, Tuesday 28 November 2006 at about 2 pm.

TR’s vehicle was a long wheelbase five door Land Rover Defender, about eighteen months old and in good condition. He had owned it for less than a month but was used to such a vehicle.

There was damage to the offside (east) edge of the tarmac, both outside and (in places) over and inside the continuous white edge line. This is common enough where the outer edge of the tarmac meets a soft grass verge, and it is caused by heavy vehicles running along or over the tarmac’s edge with their tyres. It is known to highways engineers and maintenance people as “overriding” damage. The effect, whilst understandably referred to as a ‘pothole’, is perhaps better described as a rut or sunken part at the broken edge of the carriageway, with, to its right, a muddy edge of the grass verge standing proud of the bottom of the rutted area.

When inspected the next day, the area of overriding damage was agreed to be 51 metres long. Although there was a clear possibility that heavy emergency vehicles attending the accident, and ordinary traffic passing the obstruction caused by it, might well have exacerbated the damage, the judge made findings as to the state of the road at the time of the accident, after hearing the evidence of two accident reconstruction engineers, much of which was agreed. She found on the balance of probabilities that at its widest the damaged area extended about 150mm (just under 6 inches) inside the edge of the continuous white edge line and, in places, obscured or obliterated that line: [18]. That widest point was about 15 metres from the beginning of the 51 metre length of damage [13]. She found that the deepest part of the damaged area was something more than 80 mm (3 inches) deep, but by how much it was not possible to say: [22]. The Land Rover tyres were 235 mm (9.5 inches) wide: [24]. She found that TR must have put his offside tyres into the damaged area, that he entered it at a point before the place at which it intruded most into the carriageway [28], and that at the point where he entered it the continuous white edge line was either missing or difficult to see [39].

The evidence of both TR and a driver travelling behind him was that the Land Rover was more or less past the overtaken car by the time anything went wrong. TR described an unexpected bump and his steering becoming very uneven. One or other of his passengers at some stage said, or had said, something such as “look out for the potholes”. He thought he was in command of the situation and steered to the left but lost control. The driver behind described the Land Rover as entering the damaged area and being apparently “spat out” to the left. The vehicle crossed the road, leaving a tyre mark on the oncoming southbound carriageway, and crashed into the ditch and trees off the road to the nearside (west). Further reconstruction was made difficult by the gravity of the accident, particularly because of the likely additional damage done by the ambulances and fire engines and because the Land Rover was badly damaged, having, moreover, had to have its roof cut off to extract the injured passengers.

As is well known, the duty of the highway authority evolved from the common law duty of the local inhabitants to maintain a highway. That duty was an absolute one but was enforceable only by what today would be called a public law action. The inhabitants were not liable for any private loss caused by the state of the road, unless there was misfeasance, as distinct from nonfeasance or simple failure to maintain. That origin is the explanation for the manner in which the law has been expressed since the Highways Acts 1959 and 1961, and in which it now appears in the Highways Act 1980. The duty to maintain contained in section 41 (which includes repair – see section 329(1)) is now broken by mere nonfeasance. It remains an absolute duty, and it may be invoked not only in relation to civil claims against the authority arising out of an accident but also by action to enforce the maintenance of the road (see section 56). It is a duty to put and keep the highway in such a state that it does not entail danger to those who use it in the manner ordinarily to be expected: see Rider v Rider [1973] QB 505 at 514D-H. However, so far as civil claims arising out of accidents arising from non-maintenance are concerned, the liability of the authority is limited by section 58 to the case where it has not taken such care as was in all the circumstances reasonably required to render the highway not dangerous to traffic. The onus of proving that all reasonable care was taken lies on the authority.

The issues which called for decision were therefore these.

Was there a breach of s 41, i.e. was the road in a condition which exposed to danger those using it in the ordinary way? If yes:

Was the accident caused by that breach? If yes:

Had Devon made out the statutory defence under section 58 of that Act, i.e. of taking all reasonable care? If no:

Was there any contributory negligence on the part of TR?

The Judge’s finding that the road was, in places at least, dangerous is implicit in several passages of her judgment, for example paragraphs [106], [110] and [112]. She was plainly entitled so to hold and there was little doubt about it on the facts. As a matter of abstract logic, I am content to accept the proposition of Lord Faulks QC for Devon that the fact that its highways staff would have regarded the defective road edge as a category 1 defect according to their working manual, and thus as requiring immediate attention on grounds of safety, does not of itself mean that the road was dangerous. It might theoretically be possible to postulate a case in which a defect met the manual’s definition of a category 1 defect but yet there was no prospect of anyone being put in danger from it in the course of ordinary road use. However, it is not easy to imagine such a factual scenario and on the facts of this case the reason why the defect was category 1 was quite plainly because drivers might be put in danger by being thrown off course.

Devon’s first ground of appeal is expressed as follows:

The Judge was persuaded, at a hearing some weeks after giving judgment, to grant permission on this ground. In doing so she said this:

Lord Faulks’ submission is, as it was at trial, that the facts as found by the judge were analogous to the position in James v Preseli Pembrokeshire DC [1993] PIQR P114. That was a tripping case in which there was evidence that an extensive paved area was generally in poor condition and in places dangerous, but the points at which the two claimants had (separately) tripped were identified and involved very limited unevenness which could not properly be called dangerous. This court held, inevitably, that in order to succeed in a claim made on the basis of a breach of section 41 it is not enough to show that the highway was dangerous somewhere other than where the accident happened; it must be shown that it was dangerous where the accident occurred, and that the accident occurred as a result of the danger.

The present case is not analogous at all. On the facts found by the judge, TR crossed into the rutted area somewhere before the point of maximum intrusion of the rut into the carriageway, that is to say somewhere before the widest part of the rut. But because he was travelling forwards at (so the judge held) about 45 mph, he promptly passed into the widest area of the rut, and that, on any view, was a dangerous part of the road. On the measurements set out by the judge, that could not have been less than 15 metres after he entered the damaged area, and probably less. The judge’s clear finding, which was inevitable on the evidence, was that it was the rut which caused him to lose control and thus caused the accident. That is enough to demonstrate a breach of section 41 which caused the injuries.

Moreover, it seems to me reasonably clear that the judge in fact found that the place described as the ‘point of entry’ into the rut, was in any event itself dangerous. It is true that at paragraph [105] she accepted counsel’s submission “that the overriding damage in sections of the road which were not adjacent to the deep pothole did not of itself render the highway dangerous.” That appears to be no more than an acceptance of the proposition that not every piece of crumbling road edge is, in the absence of a rut formed of a kind which may throw a vehicle off track, dangerous. In the very next paragraph the judge held, speaking of the evidence of the road engineer called for TR, that:

By the time of this appeal, the principal issue on the statutory defence was whether Devon inspected the road sufficiently often. Even at trial, there was no complaint about its classification of defects, nor about its speed of response to their discovery, nor about the appropriateness of its remedial action. At trial, TR had additionally contended that the road ought to have had reconstructed or reinforced edging (“haunching”) provided, but the judge rejected that contention and there is no challenge to her decision. As to inspection, Devon’s settled practice for this road and a great many others of the same kind in the county was to inspect at six monthly intervals. The judge found that this was not enough to discharge the onus laid upon Devon by section 58 to show that it had taken such care as was in all the circumstances reasonably required to secure that the part of the highway to which the action relates was not dangerous to traffic.

There was and is a non-statutory code of practice: “Well maintained highways”. A 2001 edition had been produced by an ad hoc group of highways officers from various authorities, sponsored by an equally ad hoc but significant group of organisations such as Transport for London, the Local Government Association and the Technical Advisors Group, representing unitary metropolitan authorities. A 2005 edition was produced by a similar ad hoc group of very experienced highways officers, now entitled the Roads Liaison Group and convened as a specialist member group of the Chartered Institution of Highways and Transportation. One of a great many areas of practice which is traversed by these codes is frequency of inspection. Introductory paragraphs, numbered 9.4.5 and 9.4.6 in the 2001 edition, and somewhat expanded in paragraphs 9.4.5 to 9.4.8 of the 2005 edition in terms which incorporate the expression ‘risk assessment’, suggest that the regime for frequency should be based on consideration of listed factors; they include category of road, traffic use, incident and inspection history. This introduction contains a statement that although category of road and traffic use will be the main determinants, the other factors should be considered, together with an on-site ‘reality check’. The codes then proceed by grouping roads according to size and type into five categories, with strategic routes at the top and local access roads at the bottom. A frequency for inspection is then suggested, which is monthly for the first three categories, three monthly for the fourth, and annually for the fifth. The third category is described as “Secondary Distributor” roads. Which highway goes into which description and category is left to the authority. Devon treated the road in the present case as a secondary (or local) distributor road and so the suggested inspection frequency according to the code is one month.

Devon had its own manual for road inspection and the treatment of defects. The edition before the judge was dated March 2004, after the 2001 code but before the later 2005 edition. Devon’s manual divided highways not into the five categories employed by the code, but into ten, with national primary routes at the top and mere tracks or green lanes at the bottom. Just as in the code, a separate frequency of inspection is prescribed by the Devon manual for each category. The top two categories call for monthly inspections, the next three for six monthly ones, the three after that for annual inspections, whilst the tracks and “minor lanes” do not require any specified periodic inspection. The present road falls, under this classification, into the fourth category down the list (albeit actually labelled ‘category 6’), namely of “Local Distributor” roads, and with an inspection frequency of six months.

The judge concluded that Devon had failed to make out the statutory defence because it had not justified its departure from the recommended inspection interval referred to in the code. She held at [121] that six monthly inspection was insufficient, both generally (for all its ‘group 6’ roads) and specifically in relation to this particular road. Later parts of her judgment make clear that her finding was not simply that six monthly inspection was insufficient, but that only monthly inspection as per the code would discharge the duty to take reasonable care (see [123] and [132]).

In arriving at her conclusion upon the statutory defence the judge said this at [115] and [116]:

In this approach, the judge fell into error. Despite the recognition in the opening words that the code was non-mandatory, this approach amounted to treating it as a mandatory standard which had to be adhered to unless there was a positive reason to depart from it. Whilst the code is clearly evidence of general good practice, its status must not be overstated. It has no statutory basis and its own terms are explicit in a section carefully entitled “Status of the Code” (this quotation is taken from the 2005 Code; the 2001 Code contains an almost exactly similar passage):

The evidence before the judge reinforced what the code itself says about its status. It showed that several other local authorities adopted inspection intervals different from those set out in the code. Specific evidence was adduced from Kent, Cumbria, Tameside and Surrey, presumably selected as a spread of chiefly rural, urban or mixed areas. Those authorities varied significantly both from the code and as between each other in their treatment of inspections. Kent used the same five groupings or categories as does the code, but adopted different intervals, some more frequent and others, including for secondary distributor roads, less frequent. It attributed roads to the categories on a basis different from the code, relying on HGV traffic rather than overall use. Cumbria used only four groupings with intervals which varied from the code in the second and third. Tameside used the code’s five groupings but added a sixth for town centres; all its intervals were longer than the code’s except the town centre group. Surrey had only three groupings; all its intervals were longer than those of the code except for the bottom grouping of local access roads, where it was the same. All four of these authorities had arrived, by these different routes, at settled policies for the inspection of local distributor roads at six monthly intervals, just as had Devon. In addition, one of the highway maintenance experts, Mr Luck, called by TR, gave evidence that his own informal and not specially extensive researches had turned up six more authorities which inspected local distributor roads at intervals less frequent than the code suggested; of these, four used the same six monthly interval. He knew of six other authorities which, by contrast, exactly followed the code categories and intervals. Further, again consistently with the status of the code, it is the plain fact that the categorisation of roads is a matter of assessment left to the authority rather than of hard and fast rule; that is illustrated by Kent’s slightly different criteria and by the fact that the road here in question, when it crosses the county boundary into Somerset, is treated as a category 4 link road so that the inspection frequency suggested by the code becomes three monthly.

The judge dealt briefly at [118] with the evidence of other authorities with six month inspection intervals. She held that it was “of little or no assistance in this case”. She said that that was because she did not, except in the case of Kent, have evidence of the reasons why the other authorities had departed from the code. That might be a perfectly legitimate conclusion if it were once correct that the code provides a mandatory norm from which departure must be justified by reasons given, but it does not. At the very least, the evidence of the practice of other authorities pointed towards a respectably held view, amongst professionals charged with highways maintenance, that six monthly inspections of local distributor roads were a reasonable response to the duty to maintain. On the well understood Bolam principle (Bolam v Friern Hospital Management Committee [1957] 1 W.L.R. 582) that evidence went towards showing that Devon had exercised reasonable care in its general policy for such roads. The judge appears to have regarded as relevant the fact that Somerset inspected this particular road, further on to the north, at three monthly intervals; if that was relevant, so was the evidence of the practice of other authorities, indeed more so since it provided a measure of consensus.

It is certainly true that the evidence did not disclose what if any discussions or, as the judge put it “study” (see [19] above), had preceded the adoption by Devon of its inspection intervals. It did, however, disclose an exercise which Devon had undertaken in 2004 comparing its road maintenance manual with the then 2001 code of practice. The comparison was not limited to inspection intervals but traversed the whole of the code. As to inspection intervals it recorded the bare statement that the local practice complied with the introductory paragraphs 9.4.5 and 9.4.6. It then stated, as to the specific intervals recommended in that code at 9.4.7:

Mr Wilmington had also referred in his statement to the fact that Government funding of local authorities proceeded on the basis that highways spending needed to be considerably less in areas of sparse population, such as Devon, compared with the national average. This was taken by Mr Luck, the highways engineer who was called by TR, to be a plea that monthly inspection of local distributor roads was too expensive to be justified. In fact, Mr Wilmington’s statement did not go this far, but rather simply recorded that the difference in cost between monthly and six monthly inspections would be £23000. Since that was the total cost for the whole county, it is unsurprising that he accepted in oral evidence that it was readily affordable, although he did say that an authority had difficult choices to make on the basis of limited resources. The judge seems to have thought, at [59], that Devon was advancing the cost of £23000 as a justification for departure from the code, which would certainly have been scant basis for it, if departure had to be justified. In fact it seems that it was not advancing this argument, but in any event this is a further example of the erroneous approach explained above.

For these reasons, the judge’s finding that Devon’s adoption of an inspection frequency of six months for local distributor roads generally was a want of reasonable care cannot stand. It was founded on an erroneous approach to the code.

However, the judge also held that this particular road called for greater frequency of inspection. Although that finding was in part predicated upon her flawed findings about the policy for local distributor roads generally, the evidence before her was by no means confined to the general. It extended to the detailed inspection reports for this road, over a period of the three and a half years before this accident. These were the subject of detailed examination in oral evidence, during which the judge described them as an important piece of evidence. They demonstrated that on every inspection except one, the last, the road had been found to have significant safety defects calling for immediate or urgent repair. A typical inspection in June 2003 revealed some 18 potholes and 13 pieces of overriding damage, all categorised by Devon as category one defects, thus calling for urgent attention. Subsequent inspections revealed that despite repair, similar defects reappeared every time. The only exception was the last inspection on 25 July 2006, when there were no defects, but that was because the road had only just been patched and surface-dressed on 1 June. Moreover, the evidence demonstrated that this road was known to be subject to overriding damage of the kind which was involved in the present accident. Mr Tucker, Devon’s highways officer responsible for the repair of this road, accepted in evidence that it was frequented by heavy goods vehicles and by large agricultural vehicles, and that given the sloping soft verge it was prone to overriding damage from them. The harvest months of August and September were likely to cause movement of large agricultural vehicles and would have come after the last inspection and not long before TR’s accident. It was, he accepted, not just foreseeable, but probable, that verge rutting would occur on this road which, if not dealt with, would lead to edge deterioration. When asked if, in order to keep on top of “this deteriorating road”, it needed inspection more often than six monthly, he replied “That is a possibility”. The judge also had the evidence of the driver who had been following TR on the day of the accident and who used the road regularly. He described it as getting worse and worse. Whilst correctly recording that there had been no prior accidents, the judge referred to this evidence in the course of her judgment at [49], [124] and [126]. She concluded at [128] that Devon could reasonably have been expected to know before the accident of the poor condition of the offside of this road.

Despite the flawed approach to the policy of Devon in relation to local distributor roads generally, it is clear that there was sufficient evidence to justify the judge’s conclusion that this particular road needed inspection at shorter intervals than six monthly, and I see no warrant for interfering with it. I would accordingly uphold her finding that the statutory defence had not been made out, whilst rejecting her wider basis that Devon’s policy in relation to local distributor roads generally betrayed a lack of reasonable care because it amounted to a departure, without sufficient reason explained, from the non-statutory recommendations contained in ‘Well maintained highways’.

In those circumstances, there is no occasion to investigate the alternative basis for Devon’s liability which is asserted by TR in his Respondent’s notice, namely that it had negligently put a centre line on a carriageway of less than 5.5 metres in width, which the Traffic Signs Manual advises against. The judge was, in any event, plainly entitled to come to the conclusion she did that there was no basis for saying that the presence of a white line had any bearing on the fact that TR went into the damaged area of road.

There has been no challenge to the judge’s finding of primary fact that TR was, perhaps rather unusually, overtaking on this limited stretch of road at only 5-10 mph more than the vehicle he was passing. On that finding, there was no arguably excessive speed; indeed there would not have been at a rather greater speed than the 45 mph which the judge found.

There was greater substance in Devon’s complaint that TR over-corrected on getting his offside tyre(s) out of the rut and thus ran off the road on the other, left-hand, side. The accident investigation experts both came to this conclusion. However, they also pointed to the fact that he had only about 0.7 of a second in which to react. The judge was, as it seems to me, entitled to come to the conclusion that even if he might have avoided the oversteer, this did not amount to want of reasonable care on his part.

The same, however, does not apply to the conclusion that there was no contributory negligence in entering the rut in the first place. Even if it had water in it, it was there to be seen. Even if TR had not known the road, which he did, the contours were obvious, with a sloping soft verge on the right. The judge found that at the point of entry, never mind just a little further ahead where the driver ought to have been looking, the white edge line was either missing or difficult to see. It can only have been inadvertence on the part of TR that he did not see the defect in the road and avoid it. There was plenty of room. The offside carriageway was between 2.30 and 2.35 metres wide, measuring from the inside of the edge line. There was no question of the Vauxhall which was being overtaken being driven on the crown of the road, or swerving, or doing anything which restricted the availability of the whole of that space to TR’s Land Rover. The track width of that vehicle, from outside tyre to outside tyre, was 1.72 metres. Even allowing for an intrusion into the carriageway by the damaged area to the extent of 150 mm which the judge found, there was more than ample space. TR in his own evidence was understandably less than entirely sure whether he had crossed onto the edge line. He thought he had not, but accepted that it would have been a miscalculation if he had. The judge at [133] addressed crossing the edge line but not the omission to be alert to the risk and to see the damage. On the evidence, and the judge’s findings of fact, there was simply no escape from the conclusion that TR made an error in not seeing the damaged area and avoiding it. I reach this conclusion without bringing into account the fact that TR had previously formally admitted liability to his passengers, although I have not the slightest doubt that that was an entirely realistic course to have taken. Although the error may have been one which many might make, it amounted to a significant failure to keep a proper lookout and to manage the car correctly; it had terrible consequences. In my view the only proper finding was that there was contributory negligence to the extent of 50%.

For these reasons, I would allow Devon’s appeal to the extent of substituting a finding for TR subject to 50% contributory negligence.

I agree.

I also agree.