The publisher is having computer problems so we will temporize with the proposal which was sent to WoW.
[I hope you can rotate AMPS about 30deg]
"Up to that point I thought that the 1971spike in the fatality rate for Switzerland was another error in the data. When I saw that it corresponded exactly with a large M1 spike, and that three or four other ‘data errors’ in the smaller economies also coincided with M1 peaks, my first thought was , ‘If most of the population is rushing frantically around trying to catch money falling from the sky, perhaps we shouldn’t find it surprising that we bump into each other with horrifying regularity.’
Al Gullon, of ACEs, Ottawa, Canada, was preparing a paper which he presented in Detroit in ‘97. (SAE Paper #970280 Autobahn Fatality Rates and General Speed Limits: What Really Happened in 1973/74?)
From mid-April to mid-June of ’96 he had driven over 10,000km, on the expressways of six central European countries, gathering data on traffic fatalities and discussing traffic safety with various professionals, including stops to chat with traffic officers at ten police stations along the route.
His analysis of the data, "points to only one conclusion: both the major reductions in autobahn fatality rates throughout Europe in the ‘72 to ‘74 period, and the horrific increases during the decade previous to that, can be adequately explained by the business investment activity (or lack thereof) in the sequential phases of the business cycle. The general speed limits imposed by most countries (in late ‘73) thus played an insignificant role in reducing accidents beyond that achieved by other safety measures taken in that time period."
Since that time Al has continued the research, with paper presentations in Paris in ’98 and Seoul, Korea last June. As reported elsewhere in this issue, BMW has recognized the validity of his work and its conclusions and invited him to be the featured speaker in a showcase’ of BMW’s technology during recent (6-9 November) ITS Congress in Torino.
Al will be summarizing the more interesting results of his research into automobile safety (cellphones, the young male driver) over the last four years, for WoW readers over the next few months.
END OF INTRO ARTICLE
Of Cycles, Deaths and Politicians
[That’s business cycles, traffic deaths and we’ll leave you to supply the adjective for the politicians.]
The story behind my tour through a thicket of economic acronyms (more economics than I ever really wanted to know) actually started in the ‘50’s and ‘60’s.
When I was growing up in Nova Scotia I don’t remember any speed limits outside of built-up areas but I do remember that we had many RCMP officers as family friends .. hunting and fishing buddies of my father .. or maybe in the same gun club.
But, 40 years on, what does my son remember?
He remembers me driving down the highway on the lookout for the "enemy"(or maybe, on the annual trip to the family cottage, sitting alongside Quebec’s A20 with a half-dozen other cars fished out of Sunday afternoon traffic by a massive QPP radar trap - there could have been two-dozen stopped cars of course but they had only two ‘ticket-writing’ cruisers).
Now, so we don’t start this discussion with a misunderstanding, you should know that I don’t mind specific speed limits (SSL’s) on obviously dangerous stretchs of roadway. (The next article will talk about some great work the Germans are doing with SSL’s.) But, recognizing this shift in my family’s relationship with police made me realize that, nowadays, the only contact that 99% of the citizenry will have with police officers is when they are being given a speeding ticket for doing 15 over ...on a straight stretch of superhighway...on a sunny summer Sunday !...in light traffic!! (Oops. Sorry ‘bout that. My voice did get a bit screechy toward the end there. I’ll turn down the treble.)
It is thus easy to see that the widespread use of general speed limits (GSL’s) over the past thirty years, particularly on 4-lane highways (which, when you look at the numbers, are roughly ten times safer than regular roads) has poisoned the relationship between citizens and police ... and the poor traffic officer, wasting his/her training and talents staring at a radar screen, is as much a victim as you and I.
In the ‘60’s I was a transport officer in the RCAF, including four years in France and Germany. They gave us intensive training in accident analysis, following a program developed by the British police, as I recall.
Great emphasis was placed on ‘the causal chain’ which starts with the impact and works backward, seeking the cause of each cause: primary - couldn’t stop in time; secondary - rusty brake line failed; tertiary - poor maintenance.
The secondary and tertiary causes are rarely related to vehicle failure however. Most often the secondary will be, ‘reacted too late’ with the tertiary being, ‘preoccupied by earlier fight with spouse’ or ‘distracted by Greek God/Goddess on sidewalk’ or ‘planning next move in office politics’.
The contradiction between this training, which pointed to inattention as the major cause of accidents, and the "Speed Kills" approach to traffic safety adopted by most governments, kept pushing me to investigate the base data on accident trends at the earliest opportunity.
It came a bit earlier than expected. The recycling technology program I was running for Environment Canada was suddenly cut last February as part of deficit reduction. After several milliseconds of reflection I decided to take the cashout and start the investigation.
Correlations and Causation
The defining moment for GSL’s came when Europe joined America. In late ‘73 Germany, France, Austria, and Switzerland, among others, responded to rapidly rising road fatalities with general speed limits (GSL’s) on all roads, including autobahns*. Italy abstained from a GSL on the autostrada* until the late ‘80’s. In March, ‘74, after a heated public debate, Germany removed their GSL on the Autobahns* in favour of a ‘government-recommended’ speed limit. (*autoroutes, interstates, Autobahns, expressways, 4-lane highways, autostrada, we’ll call ‘em all expressways from here on)
In ‘74 the annual death rate obligingly dropped 30 to 40% in all countries and, in most countries, continued down in ‘75 (before stablizing around a long term downward trend, thanks to the active and passive safety technology being built into new cars then, and since). Public authorities, on both sides of the Atlantic, ignoring the recession, the oil crisis and Germany’s excellent safety record, credited this happy result to the imposition of the limits and continue even today to chant the mantra that GSL’s save lives.
This then set the first target for the investigation: find out what really happened in ‘73-’74 on the expressways of Europe.
People often impute causality to correlation of two events, A and B. The scientific process, however, begins with the identification of a possible mechanism (a theory or hypothesis) by which A could produce B. It then proceeds to test the theory, with experiments (in the ‘hard’ sciences) or epidemiological studies - aka ‘correlations’ (in the ‘soft’ ones) - before concluding that even a strong correlation is evidence of causality. A good mathematical correlation could mean only that A and B are both caused by an as yet unobserved, or at least unexplored, third parameter. (Sometimes the serendipidous sighting of such seemingly spurious siblings is the only way to ‘see’ an underlying but unmeasureable cause.)
The correlations were therefore investigated in accordance with what I call the 3P (Plausible Physical Path) principle. For example I first thought that the ‘baby boom’, that is, when the boomers hit their late-teens and early 20’s, would explain the rise and fall of the death rate. The timing was about right for North America but, when I put the numbers into the laptop, the graph showed a valley where I expected a peak. The baby-boom happened almost ten years later in Europe.
It next occurred to me that a big drop in alcohol consumption during the recession might explain the drop in the annual expressway death rate. When I got the appropriate statistics into the computer the graph showed some countries up and some down, and by only a small percentage either way. Scratch the alcohol theory as an explanation. (Of the observed variation that is. Alcohol was and, in spite of some improvement, unfortunately remains a major cause of traffic deaths.)
I then turned to economic parameters and, I must admit, sort of stumbled over the M1 connection along the way. While trying to find out why the same-year correlations of GDP changes with fatality rate changes were so poor I noticed that the changes in the fatality rate seemed to predict, and very well, the GDP in the following year.
The realization that GDP increases in year X results from the ‘gleam in the entrepreneurs eye’ in year X-1 led me quickly to M1 ...and to the interest rate ...and to new car registrations (for potential investors it answers the question, ‘Are consumers in a buying mood?’) In fact combining several such ‘economic indicators’ into a single ‘Drivers’ Economic Distraction Indicator’ (DEDI -pronounced ‘dead-eye’) seems to produce a better correlation than any of the indicators taken alone.
M1 and Traffic Deaths: Not-so Spurious Siblings
Those chats with police officers in five different countries (really intelligent people living with the problem on a day-to-day basis) provided the key to understanding the connection between economics and accidents.
I had expected that each officer would have a pet theory on what causes accidents. However, when we got around to the main question, ‘Based on your 20-odd years of experience what, in your personal opinion, is the main cause of accidents on the expressway?’ , every officer gave the same response, ... too little concentration on the driving task.
Lots of M1, when paired (as it usually is) with low interest rates, tends to get the entrepreneurs among us excited enough to forget lots of things, including that they are supposed to be driving a car. This high level of ‘economic distraction’ will not, however, be limited to the ‘money-men’ and senior management. The sense of urgency, the excitement of a business expansion and the pressure to drive equipment (and people) to design capacity and beyond will permeate the whole company, right down to the shop floor (and out into the delivery truck) and will be riding home with every employee (along with the dreams of the promotions to be had in the new plant).
The extra effort, and unpaid over-time, in the planning and construction stages of increasing production will be invisible to the GDP statistician. Hence M1 spikes will affect GDP with ‘long and variable lags’ (Dr. Milton Friedman’s phrase) but will immediately drive the entire workforce to distraction ... and into accidents.
When I saw that a huge rise in fatalities in Switzerland in 1971 corresponded exactly with a large M1 spike, and that three or four other fatality peaks in the other countries also coincided with M1 peaks, my first thought was , ‘If most of the population is rushing frantically around trying to catch money falling from the sky, perhaps we shouldn’t find it surprising that we bump into each other with horrifying regularity.’
Moreover, the reaction of the business community in each country to their first experience with stagflation can be ‘read’ from the interaction of their respective expressway fatality rates with the money supply. (Fig.5b in the paper) Since it is generally thought, within the economics field, that such M1 peaks result from an error (of commission or at least omission) on the part of financial bureaucrats one is led naturally to a second "outrageous observation".
‘Dentistscap theirs, surgeons bury theirs but government economists (and/or their political bosses) see their mistakes recorded in blood on the nation’s highways. [Another reason to call it ... ‘the dismal science’.]’
The Down Side
For most people I have talked to the ‘Physical Path’ between ‘too little concentration on the driving task’ and job-related stress in an over-heated economy, is only too ‘Plausible’. (In this I would include euphoria as a stressor - when things are going really well we tend to have a lot of pleasant daydreams.)
But what about an economic recession when the real demand is far lower than that expected? I think it’s obvious that the pressure is off. The extra capacity put in place to meet the ephemeral demand means that the equipment and people are now operating well below capacity.
There may even be a bonus to driver alertness. During boom times we’re operating as the ‘hunter’, focusing intently on our ‘prey’. But when the economy turns sour we switch to ‘hunted’ mode, alert for possible danger from any direction.
The delivery truck driver is a good example. When business is expanding he’s thinking, "If I don’t get these extra deliveries done by six, I can kiss that promotion goodbye!" In the recession which follows he’s thinking, "Gee ... it’s really nice to have only three deliveries on this route today .... But I gotta watch out. They’re lookin’ to lay people off at the plant. If I so much as scratch this truck, I’m history!"
A Logical Conclusion
Qualitatively the case was very convincing but in trying to put an engineering handle on the data I bumped into a problem. Because of large changes in government economic policy in all countries in 1975, and also important changes in safety technology (passive and active) coming into the fleet in the mid-’70’s, I found I could only use the four years ‘71 to ‘74 for calculating correlations.
However, the high inter- and intra-country consistency of the 4-year correlations between economic parameters and the expressway fatality rates, when combined with the fact that Germany, without a general speed limit on the autobahns, now has a better safety record than those countries with a limit, led me to conclude that "... both the major reductions in expressway fatality rates throughout Europe in the ‘72 to ‘74 period, and the horrific increases during the decade previous to that, can be adequately explained by the business investment activity (or lack thereof) in the sequential phases of the business cycle. The general speed limits imposed by most countries (in late ‘73) thus played an insignificant role in reducing accidents beyond that achieved by other safety measures taken in that time period."
Next month, "Why did the Viper brake so hard and long?" and "How cum the Ferrari driver got off? among other interesting items from 10,000km of expressway travel.