Archive for October, 2006

E-voting round-up

Posted on October 30, 2006, under evoting, general.

Just in time for Halloween (maybe we should start producing E-voting machine costumes) there are some more significant developments on the Electronic Voting front.

E-voting abandoned – in part – in the Netherlands

Firstly, in the Netherlands and mirroring what happened in Ireland 2 years ago, E-voting has been abandoned in 35 districts (including Amsterdam) just 3 weeks before the general election administrators are now looking for ballot boxes.

The machines in question this time are SDU machines which are generally considered more modern than the Nedap machines, and are used in about 10% of the Netherlands. Although the Dutch anti e-voting group have still not managed to get their hands on an SDU machine their concerns triggered some additional testing which revealed that the SDU machines are vulnerable to same electronic eavesdropping problems they demonstrated with the Nedap machines.

It remains to be seen whether or not the Nedap machines will be used in 3 weeks time, but already there are negative soundings from the ministry concerned and the Dutch group had previously demonstrated that the Nedap machines can be eavesdropped. I share Rop’s belief that it is only a matter of time.

Joe McCarthy wins FoI appeals, but new concerns raised

A few weeks ago I met Joe on Grafton St. and he relayed the great news that after over 3 years, he has finally suceeded in getting access to some more of the material he requested under the Freedom of Information Act. Previously the Department of the Environment have appealed these requests on the basis that it is commercially sensitive; in order words that it would harm the business of the vendors.

Having seemingly prevailed against that spurious challenge, it now emerges that a new argument is being made; that the release of the material would facilitate the commissioning of an offence. Hacking e-voting machines is an offence in Ireland, and now that the Dutch group have demonstrated that it is possible, the Information Commissioner may take the view that further releases of information could further facilitate offence.

You can read Joe’s mail and the ensuing discussion. One excerpt from the investigators letter which I thought was particularly interesting is;

You may argue that the recent hacking of the voting machines in Holland means that the competitive position of Nedap cannot be further prejudiced through the release of the above records. However, it seems to me that the release of the records would involve the disclosure of information not already in the public domain, which could indeed further prejudice Nedap’s competitive position.

That would seem to be an astonishing revelation; what information could possibly make Nedap’s competitive position even worse?

Joe has also set-up the excellently well-named Fiasco.ie which serves as a source of information Joe has collated in relation to the E-voting and the Poolbeg Incinerator projects. As if all this wasn’t enough, Joe also appeared on Radio 1 yesterday, the stream is here and the relevant discussion is 1 hour in.

Some clarifications

Talking to various journalists over the last week has revealed two common sources of confusion which I thought I would try and lay to rest. Firstly, the exact type of attack which is causing such concern in the Netherlands right now – electronic eavesdropping – does not require physical access to the machines, but is in fact a passive form of attack. The situation is analogous to a glass ballot box which lets you see from a far how someone has voted.

The Irish machines are prone to this exact attack and it is particularly relevant to a Constitutional challenge in Ireland. There are strong precedents for challenges to ballot secrecy in Ireland and the courts have previously found measures which impacted voting-secrecy unconstitutional. Despite the Taoiseach’s rhetoric, no physical access to the voting machine is required for this attack, all that’s needed is a radio receiver, even a short-wave radio will probably do. An aerial is needed, but not a big one, there is no reason all of this could not fit in your pocket.

Although this presents a risk of voter intimidation, that someone might stand in the room and monitor how you vote, that has it’s own risks (the victim may go to the police). There are much more likely and clever forms of attack. During the last general election, I was a spoiled-ballot adjudicator in the Dublin South Central constituency. As we deliberated over various spoiled ballots, it became clear that a disproportionate number of unstamped ballot papers had Sinn Féin as a first preference. Each vote was clearly individually marked, making ballot stuffing an unlikely cause but instead the returning officer suspected that a polling officer had failed to stamp the ballots of likely Sinn Féin voters as they went to vote. An investigation was promised, but I’m not certain if much came of it.

With the Nedap system, a polling officer could similarly monitor the actual preferences of voters, and before the voter had a chance to press the cast vote button the officer could intervene or remove the power from the machine as it is pressed (in which case a vote will not be recorded).

The only realistic way to fix this problem with the Nedap machines would be retrofit a faraday cage (metallic shielding) around all of the components which emit the detectable signals. Not only is this costly, but the the actual buttons that a voter presses may be in this list of components, making it impossible. This is just one of the many reasons it is beyond ludicrous to suggest that these machines will ever be used.

The second item of confusion is surrounding the recommendations of the commission and the costs involved. Following Simon’s post several journalists got in contact with me about these particular things.

The actual recommendations concerning direct hardware modifications are R5, R6, R7, R9 and R10 which are all to be found in part 8 of the Commission on Electronic Voting’s second report. Additionally, part 5.2 (in C10) of that same report asserts that the Nedap system …

is not subject to any meaningful independent audit of its vote recording function. Thus the paper system is superior in this respect.

… which as far as we’re concerned is game over, and implies a change of hardware.

The Register screws the issue, again.

Last week, another bizarre and mis-informed article emerged on E-voting from Thomas C. Greene over at the Register. In that article Thomas linked to his previous articles on the issue but failed to link to one other in that series; Fergal Daly’s rebuttal of nearly every point he made, and Fergal wrote that over 2 years ago. You can read a series of other replies, including mine, in the ICTE thread.

Fianna Fáil Árd Fheis

And lastly, at the end of this week we’ll have the Fianna Fáil party. Between now and the end of the Árd Fheis, a further €16,000 will have been wasted. That the Taoiseach and Minister Roche are behaving irresponsibly on the issue is now clear. As Noel Whelan pointed out in his opinion piece in the Irish Times last Saturday;

Surely if the Taoiseach and his colleagues have learned anything in recent weeks, it is that the public realise that to err is human. The Government should have the decency to admit their error of judgment on electronic voting, apologise for it and trust the voters to be proportionate in their response.

The Árd Fheis is the perfect opportunity for the grass-roots of the party to hold the Taoiseach to account, but I’m betting there won’t be so much as a peep on the issue. As I said to one member of Fianna Fáil;

If the party of which you are member has no mechanism by why which such stark realities can be brought home to policy changes with the minimum of waste, then it’s simply never going to be worth respecting. If a change can’t be effected when you have objective reality on your side, it’s just a no-hoper.

I certainly won’t be holding my breath. I can’t but agree with Noel Whelan’s assessment;

Even though the Cabinet subcommittee has just begun its work, Minister for the Environment Dick Roche was confident on TV3 last weekend that the machines would be used in elections due to be held in 2009. Most voters will see this latest charade for what it is – an attempt to long-ball the ultimate decision to scrap this e-voting proposal well out past the next election in the hope that the current Government won’t be blamed for wasting the money. It won’t work and the Government should cut its political and financial losses now.

The battle against e-voting in Ireland has been won, what this is is about now is finally putting a stop to the senseless waste and beginning to lay some foundations for truly productive electoral reform. What we have right now is stomach-churning politics at its very worst.

Lethal weapon

Posted on October 30, 2006, under general.

Almost every time I drive I think of Tony Kelly, a friend, and friend of the family, who was killed when a truck hit him on the N4 at Palmerstown. There was never a suggestion that bad driving or speeding was involved, crossing a dual carriageway at a bend as a pedestrian was the major contributory factor, but the fallout from the accident still gives me pause for thought every time.

Car Crashed

Tony worked in the Cara Cheshire home for the disabled in the Phoenix park and was a great source of fun and energy there (for laughs he would phone in to the various late night Dublin radio shows on FM104 and 98FM and play devil’s advocate on some of the most ridiculous topics). Much of his spare time was spent on community work in Ballyfermot, at youth clubs, on trips (I remember great ones to Kilkenny to Kerry) and working on his new house, which he was decorating himself. Left behind after the accident was Tony’s wife and ten year old son.

A memory which will forever remain indelible in my mind is the now current Lord Mayor, Vincent Jackson, distraught in our living room, left wondering why such a thing should happen to so undeserving a person when at a same time there was no shortage of people who would hardly be missed in the area. Every time I drive around a bend, I’m conscious that there could be another Tony on the other side.

if you’re Irish, chances are you too probably know someone who has been affected by a traffic accident and if you drive, you’ll be all the more aware of it. Right now, there’s somewhat of a political and media hysteria about road deaths. It’s a powerful and emotive issue, the carnage is senseless and people genuinely want to see an end to it, but we all know that there is a strong conflict between that desire and personal behavior. Everyone agrees that something should be done, but there isn’t much in the way of a sensible debate about these issues.

The problem with hysterical debates though is that they tend to vastly over-simplify problems, and we lose track of the true complexity of what we are faced with. For nearly every measure and suggestion there are complicated side-effects that do not get as much attention. There’s an increasing possibility that we can’t see the forest for the trees, that we are losing track of the simple and effective, but costly, measures that might make a real difference while getting distracted by outcries, over-complicated inventions and entirely ineffective strategies.

How do we promote a responsible driving culture?

This question does get asked a lot, and it really is key. Safer driving will have the greatest impact on road deaths. Everyone wants driving to be convenient, we want to be able to get in a car and get from A to B efficiently. Standing back from the hysteria, when you think about the number of road journeys that must occur on a daily basis, the complexity of the road network and the unpredictable nature of other drivers, pedestrians and animals it’s actually staggering that we don’t get more road deaths.

We should take stock of that, and acknowledge that the vast majority of drivers are driving safely. So the question I often wonder is how effective can we be at introducing a cultural change in the remainder? Is a boy-racer idiot really open do the kind of societal persuasion we’re talking about? What about the drunk-driving pub-goer? what about the lunatic who overtakes 5 cars just before a bend? I’m afraid I have to share Fergal’s observations on this one. These people are dim.

That’s not to say that promoting a safer driving culture is not a good idea, it certainly is, and will help to reduce road deaths, but we need to be conscious of its limitations; it is not the answer to the in incidents we are most hysterical about. So how do we deal with these people? Well we need to keep them off our roads, or to somehow mitigate the damage they inevitably cause, which is what the rest of my points are about.

Have we thought about the social consequences of enforcement?

Nearly everyone agrees that we need better enforcement. There should be more random breath testing, there should be testing outside pubs, more speeding detection, stiffer penalties for unaccompanied provisional drivers and so on. However, we should also realise the social cost of heavy enforcement on safe drivers.

Safety is hard, it relies on keen personal judgment and vigilance, there really are no simple hard and fast rules to make us all safer. The rules of the road on the other hand are a function of what can be enforced in the everyday world. It’s no safer to drive at 49kph than it is at 51kph. In pure physics terms, if you’re doing 45kph down a 15 degree incline, the stopping distance will be about twice as long as doing 55kph up that same gentle hill. It can also often be safer to speed; if you’re being tailgated in a busy 50kph zone, some times it makes more sense to accelerate and get out of the way. Right now, several kilometers of the M50 have a speed limit of 60kph, to accommodate the widening works, but if you obeyed that speed limit you would cause chaos and havoc across both active lanes.

Of course many drivers exaggerate these logical inconsistencies, they are not all that common, but they do exist. And strict inflexible enforcement generally does little to discourage the lunatics, and a lot to cause bitterness and resentment among safe drivers. There is a real danger that people would begin to see traffic police as just an unjustified nuisance, which can do very real damage to efforts to promote a culture of safer driving.

Effective enforcement should look more like targeted and intelligent policing than it should like mass-surveillance and profiteering. If on an average 2 hour journey around the roads I can spot 4 or 5 instances of criminal driving, that would suggest that some unmarked police cars patrolling with cameras can catch about 20 a day each. That type of policing tends to have a very strong network affect too, because it could be literally anywhere.

A hidden static camera snapping everyone who goes above a certain speed on the other hand does little to target the kind of people we really need off our roads.

Why isn’t reducing the waiting list for driving tests the number one priority?

An extreme example of the danger of resentment towards enforcement is that of provisional drivers. Right now the average waiting times for a driving test are over a full year, that’s just insane. It simply cannot be reasonable to expect people to sit around on their hands for a year waiting for a test. People learn to drive because they want to use a car to get from A to B and we can’t expect them to obey a law which is so incongruous to that desire. If the Gardaí did actually start enforcing the rules with provisional drivers, things would only get worse. For a start, there’d probably be a roaring trade in fraudulent licenses.

The police force is never independent from society, and as much as the text of a law may say one thing, what can be practiced in the real world is another matter. People have an innate sense of fairness, and my sense of it is that until the waiting list is more like 4 weeks, then it’s just not reasonable to “go after” these ordinary people who find themselves in an insane situation.

Last week I listened to the Minister for Transport, Martin Cullen (who incidentally is former chief executive of the Federation of Transport Operators), being interviewed at length about road safety. He suggested that our testing system should involve multiple tests in different conditions, should be more difficult, and that regular re-testing should be considered. Taken together those things could see anywhere from a 4 to 6 fold increase in the number of tests being taken.

Greater enforcement also implies greater numbers of disqualifications, which is yet more tests for the system to cope with, and lets not forget that the population of the country and car-ownership are growing at unprecedented rates. That the driving test waiting time is a core fundamental problem for the whole road-safety issue should be obvious; we can’t have better testing and we can’t really have better enforcement until we get that down. And yet the problem itself is Kafkaesque. The minister has comprehensively failed to get the waiting time down or add any significant number of testers.

What’s more the problem is treated as if it can be solved with the temporary addition of testers. Anyone with a basic understanding of queuing theory can tell you that won’t work, the number of testers needs to be increased on a permanent basis, even more so if the number of tests per person is to increase (as the Minister suggests). Why is every article on road safety not accompanied by the facts on this issue?

Are speed limits an entirely good thing?

Speed limits are a great aid to enforcement, it’s much easier to determine if a car is driving above a certain speed than it is to make an argument about a drivers potentially dangerous behavior. But that’s not to say that they are without their drawbacks, at least as currently implemented. There is one basic and simple fact about speed-limits; checking what speed you are going takes your eyes off of the road.

The more paranoid we make drivers and the more punitive we make the penalties for speeding, the less they will keep their eyes on the road. This has been shown to be a contributory factor to certain types of road deaths, particularly cars hitting pedestrians and cyclists in urban areas.

I’m not for a minute going to argue that we should get rid of speed limits, of course we shouldn’t, but why are we not putting more thought into the practicalities of how they are implemented? Most drivers can tell roughly what speed they are going without looking, but only to within 2 and 5kph. So how do we allow for this kind of tolerance without letting the effective speed limits creep up by 5kph? Has any thought or research been put into speed limits which are based on the duration of time spent at that speed?

Checking your speed every 30 seconds is a lot safer than checking it every 3 seconds, so we could have a law which says that if you breach the speed limit by 5kph or less for more than minute you’re out. This can be enforced with marked and unmarked Garda cars on the roads, and bridge cameras which measure your average speed over a time. It also means you can legally accelerate beyond the limit for a brief period to over-take. And of course if you breach the speed limit by more than 5kph, you’re out straight away.

This is the kind of law that mirrors that innate sense of fairness we all have, it’s a reasonable thing, and it’s the kind of law we tend to respect and feel good about obeying a lot more. It lets us keep control over our own judgment while making us responsible for driving safely and culpable for not doing so.

Right now, speed limits are opaque and frequently arbitrary, at a minimum we need to demonstrate to drivers why a speed limit is a particular number on a particular road, so that at least they can feel cognisant of the danger rather than bewildered by the curiosity.

Is the cost of insurance a positive or a negative influence?

I’ve never thought that the Government were serious about reducing the costs of car insurance, and in sense, maybe they shouldn’t be. The high cost of insurance creates a barrier to entry, it acts as some kind of moderating influence on the growth in driver numbers. Were the cost of the insurance not so high for young male drivers (not that they are as much to blame as would be believed from reading the media) there would be even more of them on the roads. Maybe insurance should cost even more than it does already.

But there is a counterpoint to this argument which is this; as can be seen on RoadDeaths.ie a disproportionate number of accidents occur in rural areas. In fact Dublin is one of the safest regions in Ireland on a per-capita basis. A lot of accidents happen on rural primary or secondary roads, with 80kph and 100kph speed limits. We all know the picture; smashed up cars on a ditch and hedge lined road. Much of the cause is the idiocy of the drivers concerned, but we have to take account that the relatively low population density of rural Ireland, and the appalling lack of public transport service in those areas, makes car ownership a fact of life. It’s not a luxury.

So with that in mind, what effect does the high premiums have there? It may well mean that the higher cost of insurance causes people to spend less money on the car in the first place, which is why – despite the NCT – we still see so many poorly maintained and barely-together cars on our rural roads. I don’t know that this is the case, it may well be that the same people would just pocket the difference if insurance was lower, but I think it’s worth looking at. Getting safer cars onto our roads is a key part of improving the statistics.

Where are the safer cars?

So where are the safer cars? It’s 2006, and although we don’t have jet-packs and personal helicopters, we really should have some improvements in cars. Cars have gotten vastly more collision-safe over the years, and our chances of surviving a collision are remarkably good, but the basic user interface – the part the idiot uses – hasn’t changed in about 70 years.

Why, in a technological sense, do I need to take my eyes off of the road to know what speed I’m going? Why doesn’t every car have an over-head display? Or better yet, why aren’t there some simple buttons for the common speed limits and an audible or visual bell to alert me when I go over it? Why aren’t the lower taxes for safer cars? Is engine size really the only useful measure here? (also, bigger engine size lets you overtake more safely, although they do let you drive more speedily too).

This year we’re told we’re not going to have a giveaway budget; great. So how about it, why not link vehicle tax to car safety? now that would be a good move. Bring back to 10 year scrappage deal too, and make it permanent.

We should be wary of too much technology though, sometimes people can become distracted by complexity and see it as the answer to all of our problems. GPS or GSM based trackers with speed limiters are a good example of this; can you imagine how dangerous it would be to have a device in a car which would allow your maximum speed to be controlled remotely? The implications for security and terrorism are enormous. Generally the simpler a technological solution looks, the more likely it is to succeed. Just like beeping seat-belt warnings.

Static speed-limiters, like the ones you find in buses, are simple, but they are not effective; the top speed limit is 120kph, to enforce that with a limiter is near-pointless. Generally the roads on which you can exceed that are the safest in the country anyway.

Where are the safer roads?

Why is it that we seem to ignore the literal meaning of the words Road Safety. As much as accidents are attributable to poor judgment and human nature, roads should be designed with this in mind. I’m convinced the much better quality of roads in the United Kingdom is the main reason they have so much a better a safety record.

Of course, this can often be used as a get-out, people naturally seek to blame everything but themselves, but it should be clear from the fact that Dublin has a huge volume of traffic and yet a relatively low number of road deaths that this is a real-world effect. The sooner we get modern, well-lit, safer roads built the safer our road network will be.

How do we proceed?

A principle I hold dear is that policy should be evidence based; that laws, police action and state policy should be founded on clear material and cogent argument. Of course this isn’t always possible, there are frequently issues on which we just have to make the best guess, but I think the debate on road deaths has gone too far in the other direction, we’ve lost a sense of proportion. There’s a real danger of us accepting almost any lark because we are so keen to see something done.

Every life is priceless, but unfortunately that’s not how society works. As much was the love and worth my friend Tony generated, which I can assure you was beyond any value, society does place an economic and social worth on fixing these things. Otherwise the speed limit would be 10kph and we’d be driving around with cotton-wool padding. There is a subtle and delicate balance to how our laws work, and we should be very careful to consider the secondary impacts.

We already have illogical rules like provisional drivers being given carte blanche everywhere except the safest roads in the country. Reactionary approaches are just too dangerous for my liking, imagine how much worse it could be.

This post is part of the Road Safety drive started by Damien Blake. Whose 5 suggestions I actually happen to agree with. The photograph is from Flickr user Stevefe and is CC licensed by attribution.