How many collisions occur that never get officially reported? That fact can affect the reliability of collision statistics and what the public is told about roadway safety.

Gorski Consulting has been involved in a long term study of unreported collisions and incidents. Since 2009 a specific site in London, Ontario has been monitored and incidents of loss-of-control and evidence of collisions have been documented. This documentation has then been followed up with requests to the London City Police so that their collision data could be obtained and compared to our data.

Frame taken from video in October of 2009 showing a northbound car travelling on the wrong side of Clarke Road north of Fanshawe Park Road in London, Ontario. Markers in the northbound lane were used to examine speed and position of vehicles passing through the S-Curve.

In 2017 a series of five articles were uploaded to the Gorski Consulting website describing the details of our findings. These articles corresponded to the five years, 2011 to 2015 for which police data was available and compared. We recently made a second request to London Police and have received their additional data for the years 2016 and 2017. A sixth article in now close to being completed that discusses the findings from the 2016 data.

The trends from the previously announced analyses appear to be extending to the 2016 data. The main finding from this work is that at least 80% of collisions and loss-of-control incidents that occur at this site (Clarke Road north of Fanshawe Park Road) are not documented in the official police data.

This finding may have significant implications for a large amount of research that is based on police reported data. It is not theĀ  news that a large number of official researchers, municipal transportation agencies, and other government institutions want to read or have known. When 80% of the data regarding collision occurrence is unknown how can analyses of large data files be expected to be reliable? While it is claimed that estimates can be generated from what data is known using a variety of complex algorithms few independent experts actually know the details of these calculations to confirm that they are truly reliable.

The results from our research is inconvenient and therefore publicity of these results has gained some resistance in official circles.