Knowledge Hidden in Mountains of Data

Critics can now say that they have always known it: speed trap campaigns do not lead to a lasting change in driving behaviour. People drive more carefully during the campaign, but the day after, the effect disappears.

Speed trap campaigns are what police refer to as the monitoring of road traffic, which usually lasts for a day, systematically penalising speeding. A few days in advance, they announce the offensive and inform the public during the campaign about the dangers of speeding.

The initiators believe that they can change driving behaviour with the campaign. However, concrete proof of this has been missing for some time now. This is provided by Professor Stefan Bauernschuster, holder of the Chair of Public Economics, and his former colleague Dr. Ramona Rekers in the study “Speed Limit Enforcement and Road Safety”, which has been published in the IZA Discussion Paper series.

The team of economists evaluated millions of datasets over a period of four years. The study shows the enormous amount of research behind a supposedly banal statement. And it provides important clues as to which measures could be useful in order to actually have a lasting effect on driving behaviour. More on this later.

So much is clear from the outset: Information campaigns on the dangers of driving too fast are virtually ineffective. “The power of persuasion has no effect on speeders”, says Professor Bauernschuster. It is true that people drive more slowly on the days before the speed trap campaigns. But that is more because they have heard about the speed trap campaigns in the media, but are no longer quite sure when the speed trap campaign will take place, because: “Once the campaign is over, we don't see any change. So, speed trap campaigns have no lasting effect on road safety”, says economist Dr. Ramona Rekers.

“Nudging” – ineffective for speeders

In connection with information campaigns, researchers refer to “nudging”, a method of “nudging” people to change their behaviour in a certain way. “But this mainly works when it is actually accompanied by a gain in knowledge”, explains Professor Bauernschuster. This is apparently not the case for speeders. People seem to know that speeding causes the most accidents – and yet they accept the risk of endangering their own lives and those of others. Unless, of course, they are scared of being caught in the act.

But how can this thinking be supported by data? Professor Bauernschuster and Dr. Rekers have found four effects that suggest that it is the fear of punishment that puts the brakes on people:

• During the speed trap campaign, drivers adhere to speed limits much more often. The number of accidents decreases significantly.
• The day after the speed trap campaign, these effects disappear, although the topic is still in the media.
• When speed trap campaigns last longer, as in Bavaria, the effect remains noticeable for longer. But as soon as the speed cameras are dismantled, the effect is gone here as well.
• On motorways, where there is no speed limit and therefore no speed cameras being used, there are no effects during the campaigns.

Mostly male, no new drivers

For their study, Professor Bauernschuster and Dr. Rekers evaluated research data from the State Statistical Offices on all 1.5 million traffic accidents reported to the police in German districts in the period from 2011 to 2014. This data provided detailed information on the place, time, people involved in the accident and the cause of the accident and allowed conclusions to be drawn about the characteristics of the people who caused the accidents: These were predominantly male and none of them were new drivers. In addition, data from more than 2,400 automatic measuring stations measured the traffic volume and speed at specific points every hour.

Our study is one of a series of studies that show that speed limits are effective, i.e. that lower speeds can prevent accidents.

Professor Stefan Bauernschuster, Universität Passau

To determine whether the accompanying information campaign reached the public, the researchers evaluated a database containing 60 million press articles as well as search queries on Google and hashtags on Twitter. The online data allowed conclusions to be drawn as to whether the subject of the speed trap campaign reached the public: In fact, the numbers shot up just before and during the campaign.

Enthusiasm for working with nuanced data sets

Dr. Ramona Rekers, former staff member at the Chair of Public Economics at the University of Passau

The idea for the thesis came from Dr. Ramona Rekers, who received her doctorate in 2017 from the chair of Professor Bauernschuster and now works for the German Pension Insurance Fund. The study was part of her cumulative dissertation.

Behind this is a collection of several scientific articles on different topics: In one article, Dr. Rekers deals with extreme weather conditions in India and their effects on unborn children as a result of malnutrition in pregnant women. In another paper, she examines the study according to which first-borns are statistically more successful in later life than second and third born babies. Using Danish registry data, she examines whether this advantage is possibly already present at birth (conclusion: it is not).

Common to all papers (besides the focus on health) is the method: Dr. Rekers always conducts microeconometric studies. This means that she searches for possible evidence of a supposed causal effect in large, finely structured data sets, from which in turn findings for social or political debates can be derived.

Studies show that speed limits work

Just like the study on the speed trap campaigns: “Our study is one of a series of studies that show that speed limits are effective, i.e. that lower speeds can prevent accidents”, says Professor Bauernschuster. The two economists assume that the cost-benefit calculation of the drivers changes during the speed trap campaigns. Due to the increased risk of being flashed by a camera and penalised, the costs of driving too fast increase in relation to the benefits, such as time savings or the joy of driving fast.

“As economists, we do not want to patronise people and forbid them from doing all sorts of things”, says Professor Bauernschuster. But: If uninvolved third parties incur costs through their own behaviour, as is the case with road traffic accidents, for example in the form of accidents or air pollution, then economists see a good reason for state intervention. This could justify speed limits, but also taxes on petrol or different vehicle taxes for particularly powerful or heavy vehicles.

Text: Kathrin Haimerl