The committee would like to express their thanks to Prof. John Adams of University College London, Prof. Gerald Wilde of Queens University Victoria, Canada, Dr. Bob Davis of the Road Danger Reduction Forum and Gordon Selway of the Cyclists Campaign Network
Update on Response to Government Road Safety Strategy
Re: Proposals for imposition of fines for non-wearing of seat belts.
Mr. Robert Molloy
Minister of State at the Department of the Environment.
Dear Sir,
With regard to our previously communicated response to the Governments "Strategy for Road Safety 1998-2002", we find that as a result of new information we must update our position. At that time we raised concerns about the proposal to increase seat belt wearing rates for vehicle occupants from 55% to 85% by the introduction of on the spot fines. We have now come into further information concerning the issue of the enforcement of compulsory seat belt wearing. We find that we now have to revise our original estimate of a further 11 fatalities to pedestrians and cyclists as a result of this part of the strategy.
The British experience of mandatory seat belt wearing.
Our original estimate of 11 additional deaths had been based on the
work of Harvey and Durbin on the effects of compulsory seat belt wearing
in Britain. To recap the situation at the time; in 1981 in the run
up to the adoption of the British seat belt law, a study was published
into the effect of seat belt legislation in a number of countries (Adams
1981). This study appeared to show that countries without seat belt
legislation had achieved greater reductions in traffic fatalities than
countries that had such legislation. That same year the UK Dept.
of Transport carried out a similar study on eight European countries that
already had such laws, this study was never published but reports of its
findings are available (Hamer 1985, See also note below on the Isles Report).
This UK DoT study predicted that the introduction of seat belt legislation
would be followed by an increase in accidents such that any benefits of
the seat belt law would be negated. In fact while this report tentatively
predicted an increase in fatalities to car drivers, it also predicted an
increase of 150/yr in deaths to other road users. In any case in
early 1983 the seat belt law was passed in Britain. In 1986 Durbin
and Harvey published their report on the impact of the seat belt legislation
(reported by West-Oram 1990). Their findings were that fatalities
to front seat occupants dropped by 460/yr, fatalities to pedestrians increased
by 135/yr, fatalities to cyclists increased by 40/yr and fatalities to
rear seat occupants increased by 75/yr. In our original letter we
were assuming a similar relationship between motorist lives saved and lives
lost among other road user groups and based our prediction on the government's
own projected savings. We now realise that we were grossly in error
in making such an assumption.
It transpires that Durbin and Harvey did not take into account the effect of a simultaneous drink drive campaign. Thus, in 1983 prosecutions for drink driving went up by 31% and the number of fatalities between 10pm and 4am dropped by 23%. The drop in fatalities at other times of the day was of the order of 3%, which may have been in line with the trend before the seat belt law anyway (Adams 1995). The seat belt law did not apply to commercial/public service vehicles and the numbers of collisions between cyclists and these vehicle types actually declined during this period. In overall percentage terms fatalities to cyclists went up by 13% and fatalities to pedestrians went up by 8%. However in terms of fatalities due to impacts with vehicles covered by the seat belt law the increases were of 40% and 13% respectively (Adams 1988). The increase to rear seat passengers was of the order of 27%. These latter figures should not be taken as being a vindication of the extension of compulsory seat belt wearing to all passengers. The 1989 introduction in Britain of compulsory seat belt wearing for children under 14 in rear seats was accompanied by a disproportionate increase in deaths to such passengers (Adams 1995). Such an outcome is obviously also of extreme concern to the GCC. Whether the most likely mechanism was an increase in accident rates or severity, or perhaps some other effect such as seat belt induced injuries is not known to us at this time.
The International experience of compulsory seat belt wearing.
Australia was one of the first countries to introduce compulsory seat
belt wearing; it is reported that these laws also resulted in non car-occupant
(cyclists/pedestrians etc) casualty rates that were higher than they would
have been in the absence of the seat belt legislation (Conybeare 1980).
It is further reported that following the introduction of mandatory seat
belt wearing in the Canadian province of Newfoundland in 1982, several
effects were seen. The total number of road traffic fatalities in
Newfoundland went up by 4% in the two years subsequent to the law. It is
also reported that during this period the proportion of total fatalities
involving non-car occupants jumped from 21% to 40%. In the adjacent
province of Nova Scotia fatalities to non car occupants dropped regularly
between 1980 and 1984 but rose after the introduction of a mandatory seat
belt law in 1985 (Wilde 1988). It is also reported that in New Zealand
not only did car occupant fatalities rise after the adoption of mandatory
seat belt law, but that this was also "accompanied by a considerably sharper
rise in fatalities of other road users" (Hurst 1979 quoted in Adams 1981).
The Irish experience of compulsory seat belt wearing.
Thus far we have concentrated on the International experience of compulsory
seat belt legislation. What has been the experience of Ireland where
such legislation has been in force since 1979? There are two standard reports
available on this issue, RS 255 "The initial impact of seat belt legislation
in Ireland" (Hearne 1981) and RS 408 "Seat belt wearing rates in Ireland"
(Golden 1991). In 1979 mandatory seat belt wearing was introduced in Ireland;
in addition, there were serious fuel shortages and the interurban speed
limit was reduced from 60 mph to 55 mph. In the second half of 1978
the blood alcohol limits had been reduced and there was an associated drink
drive campaign (Mulkeen 1982). Any or all of these measures might
be expected to bring about a reduction in road deaths. Clearly it
could be expected that 1979 would become a land mark year in the history
of Irish Road safety; what then, were the actual results? Writing in 1981,
Hearne found that seat belt wearing rates in Ireland had gone from one
in four, to one in two in the space of eleven months. In terms of
effects however, he found that "the legislation has had no significant
effect in the period considered on the severity of driver injury."
If we look at the actual crash figures then a slightly different picture
emerges. There were increases in fatalities to both pedestrians and
car users in 1979. Car user fatalities went from 248 to 258 while
pedestrian fatalities went up by one from 226 to 227 (Road Accident Facts
1982). Even more disturbing is the realisation that according to the tables
given by Golden there were 570 fewer injury/fatality crashes involving
motor vehicles recorded in 1979 compared with 1978, so the fatality rate
for car users per recorded injury accident actually went up.
Hearne himself notes that "the number of drivers killed in single car crashes
was higher than had been expected". Single car crashes are almost
universally associated with inappropriate speed; that they should have
increased in severity in a country that had just reduced the general speed
limit is somewhat perplexing. What other effects might be seen? According
to the table below it will be seen that of the years between 1973 and 1982,
1979 shows the highest number of material damage accidents. In terms
of total accidents, the total for 1979 is the second highest in the period
considered but is only 50 below the peak seen in 1978.
Accident type |
|
|
|
|
|
|
|
|
|
|
Fatal |
|
|
|
|
|
|
|
|
|
|
Injury |
|
|
|
|
|
|
|
|
|
|
Material Damage |
|
|
|
|
|
|
|
|
|
|
Total |
|
|
|
|
|
|
|
|
|
20346 |
It might be tempting for some in the road safety industry to portray an increase in the proportion of material damage accidents as being a positive sign of increased safety. It should be noted however, that for a pedestrian or cyclist being struck by a car there is no such thing as a material damage accident. Any increase in such accidents clearly indicates that the roads are becoming substantially more dangerous, regardless of whether or not this is reflected in the accident statistics. It should be noted again that the best estimate for Ireland indicates that only 9% of serious injury accidents involving cyclists are officially recorded (Brennan 1998). Hearne chooses not to explore the apparent non-effect or possible negative effect of the Irish seat belt legislation. Instead it is asserted that the seat belt legislation will only become effective when the overwhelming majority of car occupants become belt users. In this context a wearing rate of 80% or greater is suggested as being effective in substantially reducing deaths and injuries. The only evidence offered in support of this assertion consists of two graphs dating from the time of the energy crisis of the early 1970s; these purport to show a relationship between Australian road death figures at that time and the wearing of seat belts. Hearne does not refer to any other countries in his search for supporting evidence. This is unusual since Sweden had introduced mandatory seat belt wearing not long previously. The Swedish mandatory seat belt legislation was introduced in 1975, and as in Ireland this was also accompanied by an anti drink drive campaign. In contrast to Ireland, Sweden achieved wearing rates of 85% or higher. Hearne's assertion notwithstanding, the Swedish legislation was followed by an increase in fatalities to car occupants.
In a subsequent study by Golden in 1991 the author claimed that it was possible to show a benefit from the introduction of mandatory seat belt wearing. We have now conducted our own analysis of this report (see attached note) and it is our view that very little reliance can be put on the authors conclusions. In fact it is possible to interpret the data presented in this study, as showing that seat belt wearers are significantly over represented in the population of drivers who have crashes which are reported to the Gardai.
At this time the GCC does not have access to the detailed accident data for 1979 and is not yet in a position to examine the issue of non-occupant deaths. It should be noted that Hearnes study chose specifically to disregard impacts with pedestrians. In any case it may not be possible to pick out any effect from among the additional influences of the drink drive campaign, speed limit measures or the fuel shortages. It is timely to note that at this time the GCC is also seeking to identify possible reasons for the 78% increase in fatalities to Irish cyclists that was seen in the two years subsequent to the “European year of road safety” in 1986.
Revised estimates of additional deaths in the Irish context.
We have now drawn up a revised set of estimates of additional deaths
that would result from the reproduction of the reported effects in other
countries. Using the Irish road death figures from 1996 as a starting
point these estimates are based on disregarding deaths due to collisions
with “non seat belt” vehicles such as pedal or motor cycles and goods vehicles.
If the Government manages to reproduce the British experience in this country,
then on a yearly basis at least 5 additional cyclist fatalities will occur
and there will be at least 10 additional pedestrian deaths. On the
other hand if a simpler Newfoundland type accident pattern were to occur
in Ireland there would be of the order of 44 additional deaths. Using
the National Road Authorities standard figure for costing road deaths indicates
a loss of between £11.7 million and £34.3 million. We
are no longer anticipating that any significant downward change in motorist
fatalities will result from this measure. In fact we now feel that
there is evidence that this measure could have the opposite effect of increasing
all road fatalities (See below for the Finnish experience).
Risk Compensation vs. Selective Recruitment.
The record of mandatory seat belt use in other countries is of a complete
failure to show anything like the previously predicted savings of lives.
The risk compensation hypothesis argues that this is because drivers adjust
their behaviour in response to a perceived change in their own personal
endangerment. Thus an increased sense of security can lead to a decreased
sense of caution such that any safety benefit is consumed as an increase
in speed or some other performance. To quote directly from one author on
risk taking behaviour "… creating the illusion of controlling traffic accident
consequences with technical features such as seat-belts, ABS, four-wheel
drive, helmets, etc. will seduce people to show more risky, thoughtless
or even careless behaviour." (Trimpop 1994)
Those researchers who reject the risk compensation hypothesis have chosen to take recourse in the selective recruitment hypothesis. This argues that the seat belt laws have not worked because those drivers who are most likely to take risks are the least likely to use seat belts (Dee 1998). Golden and Hearne argue in favour of selective recruitment but do not offer any evidence in support of such an effect in the Irish context. Dee does offer evidence of selective recruitment, but offers no proof of the accompanying assertion that seat belt use does not adversely affect expectation of crash involvement. However, this assertion ignores the existence of direct empirical evidence for a risk compensation effect among drivers who do not habitually use seat belts (Janssen 1994). Thus, the proponents of selective recruitment argue that if this risk taking section of the population can be convinced or forced to wear seat belts the benefits of seat belt use would then become apparent. In other words, if those who represent the most danger to their fellow road users can be protected from the consequences of their own actions, the roads will then become a "safer" place. To our members, who are the road users most endangered by such persons, this suggestion is patently ridiculous.
The Finnish experience
Leaving aside the possible effects of mandatory seat belt laws themselves
is there in fact any evidence regarding the actual effect of the imposition
of fines for the non-wearing of seat belts? It transpires there is a case
model available in Finland where fines for non-wearing were introduced
in 1982. To summarise briefly, a seat belt law was introduced in
Finland in 1975 and in that year road fatalities in Finland actually increased,
thus reversing a steady trend of decline that had been seen in the previous
three years (see note below on Finnish experience). The overall downward
trend was re-established in subsequent years up until 1981.
In April 1982 fines were introduced for the punishment of non seat-belt
users, as a consequence wearing rates were increased from 20% to 70% in
built up areas and from 50% to 93% in rural areas. The effect
that was seen in traffic fatalities was of a steady increase in deaths
throughout 1982 and continuing into 1983. It is reported that by
the first quarter of 1983 traffic deaths had shown an increase of 46.7%
compared with the same quarter the previous year, before the introduction
of the fines (Adams 1985). As one author has put it; "... to
compel a person to use protection from the consequences of hazardous driving,
as seat belt laws do, is to encourage hazardous driving. A fine for non-compliance
will encourage seat belt use, but the fact that the law fails to increase
people's desire to be safe encourages compensatory behaviour." (Wilde 1994)
Conclusion
In conclusion this aspect of the current road safety strategy flies
in the face of international experience. Vulnerable road users already
shoulder a disproportionate share of the impact of dangerous driving.
Pedestrians, cyclists and safe motorists are expected to share equally
with dangerous drivers, the levies required to fund the emergency services,
the fire brigade and ambulance units who must attend crashes involving
unsafe car users. We are all taxed equally to cover the cost of 24hr
monitoring of the roads for unsafe car users by the Gardai.
The public health system continues via the taxpayer to subsidise unsafe
car users and their insurers who are not required to reimburse the state
for the full cost of the medical treatments they incur (“Hospitals fight
cost limit on treating victims”-Irish Times 3/2/98). Now it would
seem that we are also to be burdened disproportionately with the actual
road deaths and injuries. This is ostensibly to be done in order
to spare those who cause these deaths from consequences of their own actions.
We are however not aware of any concrete evidence, that even among this
target group, this measure will have any overall beneficial effect.
It is our view that the proper focus of the Government Road Safety strategy
should be on measures designed to reduce risks to all road users, including
the most vulnerable and least culpable groups. We wish to reiterate our
previous call for an urgent review of this particular aspect of the current
Road Safety Campaign. The current seat belt proposals are ill advised
and amount to the imposition of needless suffering on those road users
who represent the least danger to others.
Yours Sincerely
Miriam Sheerin
Shane Foran M.Sc.
Dip.Civ.Eng.Tech.
Road Safety Officer
Secretary
C.C., An Taoiseach, Minister for the Environment, Minister for Health, Minister for Justice, National Roads Authority, Garda Commissioner, Chief Superintendent Monaghan (Galway Division), Galway Corporation, Dublin Cycling Campaign, Dublin Transportation Office, John Bruton T.D., Rory Quinn T.D., Frank Fahey T.D., Senator Margaret Cox, Éamon Ó Cuív T.D., Pádraic McCormack T.D., Michael D. Higgins T.D., Senator Fintan Coogan, The Irish Times, The Irish Independent, The Examiner, Irish News, City Tribune, Galway Advertiser, RTE, Today F.M., Automobile Association, I.S.P.C.C., Irish Insurance Federation, Vintners Federation, Motor Schools Association,
The Isles Report
(this note added to web version in December 1998)
The Isles report "Seat belt savings: implications of european statistics" was commissioned by the UK Department of Transport in 1981 in response to Prof. John Adams paper "The efficacy of seatbelt legislation: A comparative study of road accident fatality statistics from 18 countries.". The Isles study was based on a smaller sample consisting of 8 European countries with mandatory seat belt legislation and 2 countries without such laws. The Isles report is important in that it drew distinctions between the effects of the legislation for car users and the effects for other road users. The report found that seat belt legislation "was followed by an increase in injury rates" and went so far as to predict a 2.3% increase in deaths among car users alone. In terms of injuries to car users the report found that compulsory seat belt wearing had no significant effect. At the time the Dept of Transports official prediction was for 1,000 fewer deaths and 10,000 fewer injuries following the proposed law, this was rejected by the author "the results are not compatible with the departments "1000 plus 10,000" estimates for front seat vehicle occupants". When the effects of such legislation on non car users were considered the report found that the "predominance of positive effects (increased numbers of injuries) is alarming". In fact the author tentatively predicted that deaths to other road users in the UK would climb by 150 per year in the event of compulsory seat belt wearing legislation. In terms of injuries to other road users the prediction was for a 11% increase in pedestrian injuries with injuries to other road users climbing by 12 to 13%. In the UK context the predicted numerical increase in injuries would have been 36,000 of whom 7,000 would be pedestrians. The countries covered by the study comprised Sweden, West Germany, Denmark, Spain, Belgium, Finland, the Netherlands and Norway with the UK and Italy used as controls for non seat belt countries. The report found that in every single one of the seat belt countries the effect of compulsory wearing legislation had been to increase the number of road accidents. While none on the individual increases was statistically significant the fact that increases had occurred in every case was of significance "the prior probability of obtaining eight positive results is 1 in 28, or 1 in 256".
The Finnish Experience
Although in the original text of our letter as quoted above we state that the Finnish law was introduced in 1975 we now believe that this statement is mistaken and that in 1975 widespread enforcement commenced of the seatbelt law which may have already been on the statutes for several years.
References (Isles Report)
References (Update on Response to Government Road Safety Strategy)