The Introduction of a Statutory BA C Limit of 50 m g/100 ml and its Effect on Drinking and Driving Habits and Traffic A ccidents

Transcription

1 The Intrductin f a Statutry BA C Limit f 50 m g/100 ml and its Effect n Drinking and Driving Habits and Traffic A ccidents P. C. Nrdzija INTRODUCTION Since 1 Nvember 1974, new legislatin n driving while intxicated has been in frce in the N etherlands. Its mst im prtant features are: a statutry BAC limit f 50 m g /100 ml, a higher cncentratin being punishable, maximum penalties f 5000 Dutch guilders, 5 years f licence suspensin and 3 mnths imprisnment, the BAC can be checked with a bld test, the pssibility f screening with a breath test and the pssibility f banning the driver frm driving in case f dubt r in case the driver makes ready t drive. In additin, the frmer law remains valid, whereby it is frbidden t drive when under the influence f alchl r ther substances t such an extent that the driver can n lnger be cnsidered t be capable f driving the vehicle prperly. W hen the plice suspect a driver f having been drinking, a breath test can be made with a 50 m g /100 ml breath test tube. Even if the result is negative, if there is a slight disclratin f the breath test tube the driver can be banned frm driving. If the result is psitive the driver is taken t the plice statin, where a secnd breath test is made, this time with an 80 m g /100 ml breath test tube. A negative result leads t a driving ban, while a psitive result must be fllwed by a bld test. Suspicin f drinking may be arused by the m anner f driving, during special plice cam paigns when all traffic is stpped r in case f an accident. Thus there are certain differences cmpared with the U.K. legislatin, fr example a lwer limit, a secnd breath test, the pssibility f plice campaigns and the pssibility f impsing a driving ban. The intrductin f the new legislatin was accm panied by an extensive publicity campaign, fcused n knwledge f the law with emphasis n the pssible cnsequences f driving after drinking in any way whatsever. It is difficult t assess the extent f plice supervisin since the change in the law. The num ber f cases f driving while intxicated, recrded by the plice, is given in Table I. The figure in the last quarter f 1974 is slightly higher than thse in the preceding quarters, and mre than dubled in the curse f The num ber f prsecutins fr ther criminal traffic ffences was reduced nly slightly in the same perid. This indicates that the plice acted n a wider scale against driving while intxicated. In part the prsecutins fr driving while intxicated is the result f reginal r natinal campaigns, during which all the traffic is stpped at certain pints. Such campaigns are held n weekend evenings all the year rund. The results are published in the newspapers. In 1975, vehicles were checked resulting in 867 prsecutins fr driving while intxicated, while in anther 1697 cases driving bans were impsed. Other infrm atin n enfrcement activities e.g. BAC f suspects, number f driving bans r actual penalties is nt available. a Institute fr Rad Safety Research SWOV, the Netherlands 454

2 The Effect f a BAC Limit f 50 m g /100 ml n Drinking and Driving 455 TABLE I N um ber f C rim in al Traffic Offences R egistered by the Plice 1971 Drunken Other criminal driving traffic ffences 1st quarter nd quarter rd quarter th quarter Ttal st quarter nd quarter rd quarter th quarter Ttal st quarter nd quarter rd quarter th quarter Ttal st quarter nd quarter rd quarter th quarter Ttal With such radical cuntermeasures, it is im prtant t establish t what extent they meet the bjective, that is, reduce drinking by rad users and, as a cnsequence, imprve rad safety. T establish the effect n drinking by rad users, this has t be measured during perids befre and after the intrductin f the statutry limit. RESEARCH INTO DRINKING AND DRIVING HABITS Design and Executin The research int drinking and driving habits cnsisted f a number f surveys starting in 1970 and repeated in 1971, 1973, 1974 and The design f the surveys in 1970, 1971 and 1973 is cmparable. In chsing the survey sites the aim was t btain a representative natinal sample. Fr this purpse thirty municipalities were selected with a prprtinal representatin f varius regins and degrees f urbanisatin. W ithin each municipality three sites were chsen. Fr practical reasns the ultim ate sample has certain shrtcmings. Hwever, these are n bstacle t a cmparisn between successive years. Each site is visited by a research team nce a year. In each municipality each site was visited sme time during the research perid, in an arbitrary rder, ne n Friday night, ne n Saturday night and ne n Sunday night. Care has been taken t visit a municipality nt mre than nce during a weekend. The investigatins cvered ten successive weekends in September, Octber,

3 456 P. C. Nrdzij and Nvember. The surveys were limited t drivers f passenger cars. The prgram required three research teams, each team visiting ne site per weekend night between 2200 and 0400 hurs. At the site, at intervals f abut 10 minutes, the first car is directed t a member f the team, wh gives a brief explanatin abut the survey and asks the driver t c-perate. C-peratin invlves an interview f abut 10 minutes, giving a bld sample and permitting tw breath tests. The bld samples were analysed with a gaschrmatgraph. Different breath testing devices were used. Drivers c-perate f their wn free will and the strictest secrecy cncerning persnal particulars is guaranteed. If it is evident that it wuld be better fr the driver nt t carry n driving, a cab is prvided fr him r her. If drivers refuse t c-perate, sme infrmatin is nted by the team members. At the survey site traffic vlum es were cunted during the hurs f the survey. In 1973 the survey was cut shrt after the 7th weekend due t the energy crisis. In 1974 a survey f limited scpe was carried ut. It was aimed at studying the shrt-term effect f changing the law and at gaining experience in carrying ut research under the changed cnditins. The different bjectives and the shrt time available fr preparatins necessitated sme alteratins cm pared with the practice in preceding years. In rder t determine shrt-term effects, it has t be pssible t study the results weekend by weekend. Based n the fact that there were again three teams, this meant increasing the prcessing capacity per team. This was achieved by shrtening the questinnaire and by determining the BAC by breath test nly. The investigatins were made at three weekends: the last weekend befre 1 Nvember (25, 26 and 27 Octber), tw weeks later (8, 9 and 10 Nvember) and again tw weeks later (22, 23 and 24 Nvember). The municipalities and sites were selected frm thse fr the preceding investigatins, again allwing fr different regins and degrees f urbanisatin. On accunt f the lw traffic densities late at night, the survey was restricted t 2200 t 0300 hurs (instead f 0400 hurs). Each team rem ained all night at the site. The mst im prtant changes in 1975 were: (a) the large-scale use f breath tests, perfrmed with the intxilyser, a limited number f bld samples was analysed by ADH-methd; (b) mvement f teams t anther site either between midnight and 0100 hurs r between 0100 and 0200 hurs (as a rule within the municipality), withut lsing infrmatin cmpletely n this time perid and requiring twice as many sites. The increase in the num ber f sites - frm 3 t 6 - in a municipality als m eant increasing the number f small municipalities in the sample. Thus, instead f ne municipality with 6 sites there were tw, with 6 sites tgether. These were cmparable in reginal character and degree f urbanisatin. In this way, the num ber f municipalities was increased t frty-tw. In municipalities included in earlier surveys cnsideratin was given t whether the usual site shuld be retained r replaced by thers. The schedule, i.e. the days n which the municipalities and sites were visited, had t be adjusted as well. The drivers checked, drivers refusing t c-perate and use f breath and bld tests are given in Table II. F r statistical cm parisn f differences in BAC-distributin between the years, the num ber f BAC-categries had t be limited, especially as ther variables were invlved in the cm parisn as well. The fllwing BAC-categries were used: < 2 0 mg; 20 mg - 50 mg; 50 mg mg and > 100 m g /100 ml. The analysis methd is capable f analysing cntingency tables after a crrectin f the numbers f each cell with a specific cnstant value. Such a weighted Pissn mdel (WPM), permits the testing f the main effects f variables and interactin with ther variables fr weighted numbers f drivers, and als the cntributin t these effects made by specific cmbinatins f categries within these variables. In this case the numbers f subjects are weighted n the basis f the rati f subjects t the traffic vlume per survey site per tw hur perid.

4 The Effect f a BAC Limit f 50 m g /100 ml n Drinking and Driving 457 TABLE II N um ber f Subjects, Refusal Rates and B A C Determ inatin 1970 year Number f drivers asked fr c-peratin Number and percentage f refusals 14% 13% 18% 3% 11% 3. Number and percentage f 1. fr whm cmplete infrmatin available 85% 87% 81% 90% 88% 4. Number and percentage f 3. fr whm BAC based n bld analyses 86% 81% 53% 9% 5. Number and percentage f 3. fr whm BAC based n breath analysis 14% 19% 47% 100% 91% R esults Frm 1970 t 1973 it was established that there was increasing drinking by drivers. Whereas in 1971 the psitive BAC-values were higher than in 1970 and 1973, the percentage f drinking drivers was between thse fr 1970 and The 1974 results are nt incrprated in the statistical analysis. Yet it can be stated that immediately after the change in the law n 1 N vem ber 1974, there were practically n drinking drivers n weekend nights. The last weekend befre 1 Nvember already shws a decrease in drinking as cmpared t earlier years. It is difficult t say what the pattern in 1975 wuld have been withut the 1974 change in the law. Hwever, at least the 1973 level f drinking culd have been expected. The 1975 level is in fact even slightly belw that f 1970 (see Figure 1 and Table III). % " ' <.20% W/////M % % >1.0% Figure 1 BA C distributin by year.

5 458 P. C. Nrdzij TABLE III BA C-Distributin by Time f Day and Year Time f day BAC (m g/100 ml) 1970 < /100 Ttal hurs hurs hurs Ttal hurs hurs hurs Ttal hurs hurs hurs Ttal *) Ttal W Ttal hurs hurs hurs Ttal (a>results f last weekend befre Nvember 1 (b) Results f tw weekends after Nvember 1 nly C nditins The reductin in drinking in 1975 was nt the same thrughut the night (see Figure 2 and Table IV). After midnight the reductin was smewhat greater than befre midnight. But in 1975 it can als be seen that the later at night, the higher the percentage f drinking drivers and the higher the BACs. In 1975, between 0200 and 0400 hurs 24% f all drivers still had a BAC abve 50 m g /100 ml, 11% f them higher than 100 m g /100 ml, and abut 35% had a BAC higher than 20 m g /100 ml. Sme 19% f the drivers with a BAC abve 100 m g /100 ml were fund befre midnight, 38% between midnight and 0200 hurs and 43% between 0200 and 0400 hurs (see Table IV).

6 The Effect f a BAC Limit f 50 m g /100 ml n Drinking and Driving ^ 60 CM 1 O. 40 CM CM >1.0% Is. I O cm ' % J % g 40' <.20% Figure 2 BA C distributin by year and time. On the ther hand, traffic densities were as fllws: 55% befre midnight, 33% between midnight and 0200 hurs and nly 12% between 0200 and 0400 hurs. These figures relate t Friday, Saturday and Sunday nights tgether. But, fr instance, the 43% f drivers with a BAC higher than 10 m g /100 ml between 0200 and 0400 hurs is subdivided int 19% fr

7 460 P. C. Nrdzij TABLE IV Distributin f Drivers with B A C >100 m g /100 m l by Time f Day and Day f Week: 1975 Time f day Day f week Friday Saturday Sunday Ttal hurs hurs hurs Ttal Friday night and 15% and 9% n Saturday and Sunday night respectively. One reasn fr this is that the reductin in drinking in 1975 was greater n Saturday and Sunday nights than n Friday night, as shwn in Figure 3 and Table V. It was als questined whether the drivers trip rigins, regin r degree f urbanisatin influenced the reductin in drinking in There was n indicatin f this. The relatin between these variables and drinking by drivers, where it applies t all the years, will nt be discussed further. The reductin in drinking in 1975 may be either the result f mderatin in drinking r f changes in traffic cmpsitin in ways cnnected with the drivers drinking habits. Pssible changes in the cmpsitin f car traffic in the weekend nights in 1975 as cmpared with the preceding years were investigated in cnnectin with the variables f time f day, day, trip rigin, regin and degree f urbanisatin. Hwever, where differences exist, they are small and d nt explain the reductin in drinking as a whle. TABLE V BAC-Distributin by Day f Week and Year Day f week BAC (m g/100 ml) 1970 < > 100 Ttal Friday Saturday Sunday Ttal Friday Saturday Sunday Ttal Friday Saturday Sunday Ttal Friday Saturday Sunday Ttal

8 The Effect f a BA C Limit f 50 m g /100 ml n Drinking and Driving n 80 - > < Q 60 - it >1.0% H Q CE 3 H 40 H < c 20 A B U I % % 80 - >- < Q Z Z> c <.20% Figure 3 BA C distributin by year and day. P ersnal C haracteristics There are sme differences in drinking by drivers f varius age grups. Early in the night the percentage f lder drivers wh had been drinking is lwer. Shrtly after midnight the percentage f yunger drivers wh had been drinking is smewhat higher, but later in the night the increase in drinking in the ther age grups is greater again. The data fr 1975

9 462 P. C. Nrdzij are given in Figure 4 and Table VI. O f curse, there are many drivers under 35 and, in additin, the percentage in the age grup increases the later it gets. This means that a large num ber f drinking drivers are under 35. % i S 60- icm > 1.0% cvj m m % 40- j % 100i <.20% CM O 20-0 Figure > 5 0 BA C distributin by age and time f r In 1975, amng drivers with a BAC higher than 100 m g /100 ml, fr example, 34% were under 25 (10% f these between 0200 and 0400 hurs) and anther 40% between 25 and

10 The Effect f a BAC Limit f 50 m g /100 ml n Drinking and Driving (19% f these between 0200 and 0400 hurs) (see Table VII). Otherwise the changes in drinking habits are abut the same fr all age grups since the change in the law. TABLE VI BAC-Distributin by Time f Day and Age: 1975 Time f Age BAC (mg/100 ml) day < >100 Ttal < 2 5 years years years > 5 0 years hurs " Ttal < 2 5 years years years hurs > 5 0 years Ttal < 2 5 years years years hurs > 5 0 years Ttal TABLE VII Distributin f Drivers 1975 with BA C >100 m g /100 m l by Time f Day and A ge: Time f day Age < >50 Ttal hurs hurs hurs Ttal The percentage f drinking drivers in 1975 is lwer than in the preceding years, bth fr men and wmen (see Figure 5 and Table VIII). A part frm this, the high BAC levels fell mre amng wmen. Finally in 1975, the percentage f wmen was higher (16%) than in the preceding years (1973: 9%; 1971: 8%; 1970: 8%). In 1974, after the change in the law, the percentage f wmen drivers had already increased during the hurs investigated. Therefre the increase is prbably due t the change in the law and cnnected with men leaving wmen t drive because they themselves had been drinking.

11 464 P. C. Nrdzij Figure 5 BA C distributin by year and sex.

12 The Effect f a BAC Limit f 50 m g /100 ml n Drinking and Driving 465 TABLE VIII Sex B A C -D istribu tin by Sex and Year BAC < 2 0 (m g/100 ml) > 100 Ttal 1970 Male Female Ttal Male Female Ttal Male Female Ttal Male Female Ttal ACCIDENT DATA The effect f the new legislatin shuld preferably als be determined in terms f rad accidents. It is nt pssible t make a simple calculatin frm the results f the research int drinking and driving habits. Thugh the general nature f the relatinship between accident risk and BAC is knwn, its precise nature may vary and cannt be indicated, certainly as regards cnditins in the Netherlands. Mrever, drinking and driving habits have been examined nly at nights during autum n weekends. The num ber f fatal accidents, in which the plice nted that ne f the rad users invlved had been drinking, is given in Figure 6 fr several successive years. The figures are mving ttals cvering perids f twelve mnths and presented quarterly. The number decreased by abut 130 at the end f 1975 as cmpared with ver 370 per twelve mnths prir t 1 Nvember Such recrds f drinking are f curse far frm cmplete and there is a danger f a change in the numbers being due t a different recrding practice. These data can thus be regarded as n mre than an indicatin f actual drinking in the case f accidents. Besides this, the num ber f accidents in which alchl was a cntributry factr wuld still have t be deduced frm them. Since m id-1975 the plice have als nted whether ne f the rad users invlved is suspected f driving while intxicated. But the definitin f this was widened with the change in the law, and cnsequently the numbers befre and after the change are nt cmparable. In additin, the plice certainly d nt recrd all cases f drunken-driving, and mrever drunken-driving des nt necessarily mean that this is the cause f the accident. An independent investigatin int drinking by rad users invlved in accidents has nt prved pssible in the Netherlands. Any effect f the new legislatin n rad safety will therefre have t appear frm the trend in the num ber f traffic accidents in the curse f the years, in the

13 466 P. C. Nrdzij absence f any knwledge abut drinking. Fatal accidents have been chsen because registratin f this type f accident is cmplete. The ttal num ber f fatal accidents ver a perid f twelve mnths is set ut quarterly in Figure 7. The num ber has shwn a decrease since the end f 1973, which seems t have stpped at the end f There is n sudden change since 1 Nvember 1974, and cnsequently it is very difficult t evaluate the effect f the new legislatin. Figure 6 Mving 12 mnth ttal f fatal accidents invlving alchl.

14 The Effect f a BAC Limit f 50 m g /100 ml n Drinking and Driving 467 The prbable effects f the cmpulsry wearing f mped helmets and seat-belts, intrduced n 1 February 1975 and 1 June 1975 respectively, must be allwed fr. It is even pssible that the factrs which, since the energy crisis, led t fewer fatal accidents during the whle f 1974 than in the preceding year, brught abut a still further reductin in In that event the new legislatin wuld have had hardly an effect, r nne at all, n the number f fatal accidents. It is equally pssible that any effect f the new legislatin links up with that f the factrs cnnected with the energy crisis and that by 1975 there were again factrs wrking in the ppsite directin. In rder t btain a clearer assessment f the new legislatin s effect, fatal accidents must be subdivided int accidents in which the legislatin is likely t be very effective and thse where it wuld have little effect. Figure 8 distinguishes between accidents invlving at least ne mving passenger car and ther accidents, and between day and night (frm 2200 t 0400 hurs), since the legislatin is likely t have mre effect n drivers f passenger cars and at night-time. energy crisis alchl legislatin mped helmet legislatin seat belt legislatin day (0400am-22.f passenger cars night (220 p m - 4am) Other traffic.-passenger cars ther traffic Figure 8 Mving 12 mnth ttal f fatal accidents by type f vehicle and time f day.

15 468 P. C. Nrdzij The num ber f accidents invlving mving passenger cars fell greatly in day-time frm the end f 1973 till the end f After this, there was a slight increase. After the end f 1973 there was a slight decrease in night-time and, in cntrast t day-time this cntinued rather mre strngly frm the end f 1974 till the end f The different trend in the curves fr day-time and night-time after the end f 1974 makes it prbable that the night-time decrease after the end f 1974 was due t the new legislatin. The reductin amunts t ver 100 fatal accidents, leaving abut 300 per twelve-mnth perid. As regards ther accidents, there was a slight decrease in the perid 1970 and later, while a barely perceptible increase in night-time accidents changed int a slight decrease at the end f Here again, therefre, the new legislatin seems t have been the cause f the m derate decrease by abut fifty t slightly fewer than 100 night-time accidents. A subdivisin int wrk days and weekends (weekend night refers t Friday, Saturday and Sunday nights) has been m ade fr fatal accidents invlving mving passenger cars, because mre effect can be expected during the weekend (Figure 9). This figure shws that the curves fr day-tim e accidents n wrk days and at weekends run almst parallel. - day (4.am-22.pm) night (22.0 0p m am) Figure 9 M ving 12 mnth ttal f fa ta l accidents invlving passenger cars by day and time f day.

16 The Effect f a BAC Limit f 50 m g /100 ml n Drinking and Driving 469 After the end f 1973 there is again a slight decrease fr weekend nights, which cntinues strngly frm the end f 1974 till the end f The number f fatal accidents decreased by 100 t slightly belw 200. This trend is hardly recgnisable n wrk days. Finally, nly fr fatal accidents invlving mving passenger cars at night n wrk days and at weekends, a subdivisin has been made int accidents with and withut ther radusers being invlved. It has repeatedly been fund that drinking plays a part especially in single-vehicle (passenger car) accidents. Frm Figure 10 it can be seen that the decrease after the end f 1974 in accidents invlving ther rad users is unexpectedly greater even than fr single-vehicle accidents. Figure 10 Mving 12 mnth ttal f fa ta l accidents invlving passenger cars at night.

17 470 P. C. Nrdzij Thus, the intrductin f the new legislatin n 1 Nvember 1974, prbably reduced the num ber f fatal accidents, the reductin being in the rder f 100 (r 35%) during ne year fr accidents invlving mving passenger cars during weekend nights. Related t all fatal accidents during a twelve mnth perid the reductin represents a reductin f less than 5%. This decrease cannt be expressed as a mre exact number since it is nt knwn whether withut the new legislatin there wuld have been a slight increase (just as in daytime), due t the cmbined effect f ther factrs and measures. In cmbinatin with the results f the radside surveys it can be cncluded that mst f the effect n accidents is due t lwer bld alchl levels. Hwever, infrmatin n the ttal traffic vlume during the hurs under cnsideratin which is anther explanatry factr is missing. SUM M ARY A N D CONCLUSIONS On 1 Nvember 1974 a statutry BAC-limit f 50 m g /100 ml was intrduced in the Netherlands. The effect f this has been investigated by studying the drinking and driving habits f car drivers during w eekend nights and by means f available accident data. Immediately after the intrductin f the new legislatin, there was practically n drinking-driving during weekend nights. There was als less drinking a year later than in the years preceding the new law while, withut it, an increase might even have been likely. The reductin in drinking was greater after midnight and n Saturday and Sunday nights. The percentage f wmen drivers increased during weekend nights after the change in the law. Amng wmen, nt nly the percentage f psitive BACs decreased, but high BAC-levels fell m re cm pared with thse f male drivers. Frm the available accident statistics it is nt pssible t cnclude directly hw many accidents are alchl-related. The different trend in the number f fatal accidents in day-time and at night-time after 1 Nvember 1974 makes it prbable that the reductin at night-time is due t the new legislatin. The greatest part f this reductin ccurs in the case f accidents invlving mving passenger cars during weekend nights, which fell by abut 100 during 1975 t slightly belw 200. In spite f this, a year after the new law, drinking by drivers during weekend nights is still s high that supplementary measures are necessary. Thus, after 0200 hurs, ne in every three drivers has a BAC exceeding 20 m g /100 ml, ne in every fur has a BAC exceeding the statutry limit f 50 m g /100 ml, while mre than ne in every ten has a BAC higher than 100 m g /100 ml. Furtherm re the effects may be expected t wear ff. Supplementary measures might include (a) in the field f enfrcement: infrming the plice f the prbable number f ffenders based n the results f the radside surveys, imprved screening facilities, simplified and quicker prcessing f the cases f drunken-drivers; (b) with regard t the public: infrming the public abut plice enfrcement and the results btained s far in terms f rad accidents and abut what drinking and driving behaviur is in fact permitted and the prvisin f mre public transprt services.