Sunday, 14 August 2016

Should Taiwan's Law And Polices Against Drink Driving Be Abolished?

I ask this question not to upset people for no reason, but because I think the objects and methods of public policy should not be wrapped in taboo but instead subject to public criticism. Nor I do not ask the question in order to advocate reckless behaviour. There are several problems with the current set up of legislation and policy introduced in 2013 that are not given critical attention elsewhere. I set out these problems below in four sections in the hope that it will encourage others, perhaps better placed than I am, to do the necessary research.



Data on traffic accidents across Taiwan is made publicly available by the National Police Agency. The data available in pdf download from their website runs over a twelve year period from 2003 to 2015. The revised, stricter legislation and policies on drink driving were introduced in 2013, so there is only three years of data since the change, and only nine years of data preceding it. However, the time series in which the data is presented is annual rather than monthly*. From simply eyeballing the data, one can infer two overall patterns although it would obviously be preferable to run a statistical trend analysis on monthly data rather than simply eyeballing the annual data.

The first pattern is that there has been a nine year decline in annual traffic accident fatalities from a peak of 3,140 in the year 2006 to almost half that amount in 2015: 1,696. If we assume that this decline is not normal statistical variation (which a larger data set might indicate it is), then there is a question as to what is causing this nine year decrease in traffic accident fatalities. The next question is how much can the reduction in alcohol related traffic fatalities post-2013 be attributed to these same causes that were already in effect and how much can the post-2013 reduction be attributed to the change in drink driving law and policies. This is an important and sensitive question because the answer may cast doubt on the effectiveness of the current law and policies on drink driving. The director general of the National Police Agency himself has publicly stated on more than one occasion that these policies are actually effective in reducing alcohol related traffic fatalities. Were the NPA's own monthly data for both traffic accident fatalities and specifically alcohol related traffic fatalities made available to the public, then this claim could be fact-checked. Instead we are simply asked to take the word of the director general on trust.

The second pattern that may be observed in the annual data made available by the National Police Agency is a long term increase in total recorded traffic accidents with an almost threefold rise from 120,223 in the year 2003 to 305,413 in the year 2015. This pattern may indicate either a genuine increase in traffic accidents or a mere increase in the reporting and recording of these accidents, though perhaps some combination of the two may be more likely. Yet nonetheless it seems odd to have a year on year increase in traffic accidents and, at the same time, a decline in traffic fatalities, and this odd phenomenon in the data would seem to be ripe for a good research project, and possibly even a PhD thesis. Although several explanations could be suggested to account for the increase in accidents and another set of explanations could be invoked to account for the decline in fatalities, there are few such explanations that could plausibly account for both phenomena.

One possibility might be the aging Taiwanese population. If the majority of traffic accident fatalities are those drivers in the 18-25 year age group (which the insurance industry generally recognizes as the most "risky" age group) then an aging population would mean that as more people in this age group grow older (and some of them die in traffic accidents), there are fewer younger people to replace them and thus fewer fatalities. Moreover, it may be true that older people tend to have more traffic accidents than younger people, but less serious ones. An aging population would mean a greater number of older drivers and a thus a greater number of overall accidents (with only a small percentage of these traffic accidents being fatalities). I mention this possibility as conjecture only, as I suspect there may be several causes behind both the decrease in fatalities and the overall increase in traffic accidents. Whether the aging population is one of those causes is an interesting question, and a significant one because if such causes can be elucidated by statistical research then we may find that the effectiveness of current drink driving legislation and policy is actually dwarfed by the impact of these other possible factors.


Too often a law or policy is judged good or bad by legislators, media commenters and members of the public in terms of the intentions of those who made it. The current drink driving set up is just such an example where the strict 2013 law and its' accompanying policies are lauded with wide approval because they are intended to stop innocent lives being lost on the roads. Yet this is a serious mistake. Laws should not only be judged on that basis, but also on account of how efficient they are likely to be. An inefficient law or policy not only doesn't work, but is expensive to enforce. It should be the task of investigative journalists to actually try to document whether a law is working or not, and to that end they require data, preferably publicly available data which would allow them to fact-check the claims made by the authorities. In the absence of such data it seems reasonable to infer that, as in other countries, most Taiwanese drivers involved in serious or fatal accidents whilst under the influence of alcohol tend to have exceeded the legal limit several times over**. This case reported in the Taipei Times last month seems to be typical, wherein the driver was found to have an alcohol to blood volume of 0.87mg per liter, almost six times the legal limit. That such people are not deterred by the already severe punishments they face (let alone by the risk of injuring themselves and others) suggests that the law is inefficient in that it likely has the most deterrent effect on those people whose low to moderate alcohol consumption makes them the least likely to cause an accident, whilst having the least deterrent effect on those people whose heavy alcohol consumption makes them the most likely to cause an accident.

Although the absence of publicly available data specifically on alcohol related cases is a problem in trying to evaluate the efficiency of current law and policy, it is reasonable to suppose that there will always be some non-trivial number of people who suffer from a range of psychological difficulties or moral failings that will corrupt their willingness to behave responsibly. These are the "lost cases"; they are those people, rich or poor, who are premeditatively willing to drive in a state of drunken stupor without regard for the consequences. Moreover, these are people who are likely to exceed any alcohol limit several fold or even higher. If the police are lucky, they will apprehend these people before they cause an accident - but the police are not supermen, and they cannot reasonably be expected to stop these people altogether. Suppose that - for whatever reason, which need not be anticipated here* - the number of these "lost cases" were to increase over the course of say, two or three years. It would then follow that the number of recorded drink driving offenses would also likely increase, even if current levels of enforcement remain unwavering, or indeed, even if they were made stricter and the corresponding punishments made more severe. The implications of this point should not be ignored.

Moral Panic

Whenever the Taipei Times, or any other newspaper or media outlet in Taiwan reports a drunk driving case in which another party is seriously injured or even killed, there is always the same moral outrage in the Facebook comments section attached to the article or in other commentary elsewhere. Typically, there are two types of reaction. The first is that the already strict laws and policies on drink driving must be made more severe still - indeed the director general of the National Police Agency seems to share this reaction, being in favour of a "zero tolerance" approach, one in which presumably the alcohol to blood volume ratio limit would be reduced to zero and prison sentences, rather than fines, would follow for the smallest infractions (the recent case of Antonio Chiang [江春男] seems to prefigure that possibility). The second typical reaction is that scooters be made illegal - even if the particular incident being reported on did not involve a scooter. These types of reaction ignore the limitations of law enforcement; their proponents seem to believe that the effectiveness of law enforcement is a simple linear function that can be increased indefinitely without inadvertent and unforeseen costs. What will such people demand if and when a true "zero tolerance" approach is introduced with an alcohol limit of zero and prison sentences for the tiniest infraction - and drink driving related fatalities continue to happen? It is not clear at that point how the law could be made more severe still without descending into farce, or worse.

The further this moral panic is pushed, and the more widespread it is, the more likely it is that public "trust" or "confidence" in the police, the executive branch of the government and the legislature will be undermined as it becomes obvious that they are failing to eradicate the problem. Rather, the public must understand that drink driving related fatalities - like murder and rape - is one of a number of problems in society which probably cannot be eradicated entirely, and that while it might be good to try to reduce its' incidence, we should not want to try to eradicate it entirely due to the potential costs of doing so.

Opportunity Costs

An opportunity cost is the loss of alternatives once a course of action has been chosen. In this case, the moral outrage over and consequent police focus on drink driving incidents means at least two things. The first is that the police have fewer resources to expend on other tasks and the second is that there may be a tendency within civil society to regard the policing of driver responsibility as a matter for the police rather than for civil society itself.

The first type of opportunity cost - that the police have fewer resources to expend on other tasks besides drink driving incidents - could include other areas of law enforcement. However, there are also opportunity costs within the same field of traffic safety and these can be easily inferred from a broader consideration of what the focus on drink driving is for. Like other governments elsewhere, the Taiwan government outlawed drunk driving on the grounds of public safety; drunk drivers are more likely to injure or even kill themselves and others than are sober drivers. However there are at least two other obvious conditions that may have a similar effect to alcohol, but which the law does not presently focus on much. The first is that people who drive whilst suffering from fatigue are surely also more likely to have serious accidents. Indeed, I have personally witnessed drivers in Taiwan falling asleep at the wheel, and have heard many anecdotes from Taiwanese friends and acquaintences who claim to have fallen asleep at stop lights and busy intersections. The second is that people who use their phones whilst driving may be distracted at crucial moments and are thus arguably more likely to have serious accidents. One could add to this the recent phenomenon of the "Pokemon Go!" game, which entices players to use their phones to locate virtual creatures in real environments and thereby seems to place a severe attentional demand on players. The phenomenon of drivers having accidents from attempting to play this game on their smart phones whilst driving has already occured in Taiwan and could get even worse. Of the two, I suspect that the first condition - driving under fatigue - may be just as serious as excessive alcohol consumption, and yet perhaps even a more common cause of traffic accident fatalities.

The object of the law should be to reduce the liklihood of people getting killed in traffic accidents full stop. It should not be to reduce the liklihood of people getting killed in traffic accidents due specifically to driver consumption of alcohol. For this reason it can be argued that, if there is to be law governing the psychological condition of drivers at all, it should not be narrowly focused on alcohol, or narrowly focused on phones, or fatigue or narrowly focused on any other influencing factor. Rather, such a law should be general in its focus, providing the police with the justification to arrest anyone whom they have reason to believe to be driving under some form of demonstrable cognitive impairment, irrespective of whether the cause of that impairment is alcohol, fatigue, distraction or anything else. A law focused on a more general object of cognitive impairment among drivers could be coupled to similarly severe punishments as in the current anti drunk-driving law. Such a law would work on the same deterrence premise as the current law, but it would neither ignore factors beside alcohol, nor would it be skewed toward deterring those drivers who would consume only small volumes of alcohol.

The second type of opportunity cost is less obvious and is in any case more conjectural in nature. It is the erosion of personal and social responsibility as a result of the substitution of law and policy. In perhaps 2010 or 2011, I was astonished to learn directly from a mid-level manager of a large electronics firm in Tainan that he and his colleagues would go out to a restaurant on a colleague's birthday, drink so much alcohol that they would go to the toilet and put their own fingers down their throats in an effort to force themselves to vomit in order to make more space in their stomachs for yet more alcohol, and they would eventually then drive home in a state of drunken stupor. Perhaps the revisions to law and policy in 2013 provided an effective deterrent to this kind of monstrously stupid behaviour, but that is not my purpose in mentioning this example. It occurred to me that shareholders should want the CEO to be held responsible for this kind of behaviour, not only because it potentially put innocent people's lives at risk but also because it represents a danger to the company. Were I in a position of high responsibility within a company like that, I would want to insist that all such private parties be registered with the company in advance so that I could assign an expenses account for taxis as well as send a "minder" along to monitor compliance and make sure they get in those taxis and not their own cars or scooters. I would want to insist on employee contracts that make it clear that such behaviour will not be tolerated as it represents a significant risk to the company; losing a number of highly skilled and experienced staff to an act of preventable stupidity could be disastrous, as it takes a long time to train such people and ensure their competence.

There are ample incentives for the public, whether in their capacity as company managers, friends, relatives and so on to find reasonable ways of policing responsible driving behaviour. A well functioning civil society may also be an opportunity cost of the reflex reliance on police checkpoints and "tough" legislation from the government.


In conclusion, I think I have provided ample reason to regard claims about the effectiveness of Taiwan's drink driving law and policy with concerned skepticism. It is possible that the current set up is either ineffective, or highly inefficient with a limited effectiveness. There are also possible dangers with regard to what the police may be missing, and the erosion of trust in the police and the government due to possibly having "overplayed their hand". Perhaps the thing that needs to be done most immediately is proper investigative research on the effectiveness of the current drink driving legislation and policy. However, that research effort must be complemented by additional research into other aspects of traffic safety, such as the effects of fatigue and driver distraction.

*I have tried looking for monthly data elsewhere on the NPA's website, along with specifically DUI fatality and injury data but have been unable to find it.

**However, the low alcohol tolerance among many Taiwanese may mean that this inference about the alcohol consumption of "most" dangerous Taiwanese drivers is false. I once took a female friend out for dinner who, when we were about to leave, collapsed on the floor having only drunk perhaps a quarter of a pint of beer. I took her home in a taxi.

No comments:

Post a Comment

Comment moderation is now in place, as of April 2012. Rules:

1) Be aware that your right to say what you want is circumscribed by my right of ownership here.

2) Make your comments relevant to the post to which they are attached.

3) Be careful what you presume: always be prepared to evince your point with logic and/or facts.

4) Do not transgress Blogger's rules regarding content, i.e. do not express hatred for other people on account of their ethnicity, age, gender, sexual orientation or nationality.

5) Remember that only the best are prepared to concede, and only the worst are prepared to smear.