|Hidden Speed Camera Trial
||Judge someone not by their mistakes but rather how they react to them|
||LTSA & Police Myth|
|The 1997-2000 New Zealand Midland police district hidden speed camera trial did not reduce crashes or fatalities.
||The hidden speed camera trial reduced crashes and fatalities!|
The following chart shows the number of monthly death and injury accidents since 1993 in both the Midland police district (Waikato/East Coast) and the rest of New Zealand. The lowest line shows the ratio of crashes in the Midland district to the rest of New Zealand (refer right-hand scale.)
From this data, LTSA had claimed to show that a hidden speed camera trial in the Midland district from July 1997 through to February 2000 (shaded) reduced fatal and injury crashes when compared with data from the rest of New Zealand.
The lowest data series on the chart plots the ratio of crashes in the Midland district compared with the rest of New Zealand (refer right hand scale). During the hidden speed camera trial there were relatively more crashes in the Midland area than the rest of New Zealand as shown by the peak in the trend line during that period. So the hidden speed camera trial had at best no effect and may have actually increased serious crash rates.
Furthermore, the period after December 2000 shows an increasing trend corresponding to the implementation of the additional traffic police/high visibility/rigid enforcement of speed limits. This policy has clearly failed.
Two independent analysts have confirmed my conclusions by doing their own analyses:
The above matched scaled charts show clearly the three strong features of the data.
Firstly there were different trends in timing and strength occuring prior to the Trial. Secondly, the trial had an adverse rather than a beneficial effect where it was conducted in the Midland area. Thirdly, the introduction of a high visibility rigid enforcement of speed limits regime with greatly increased road patrols coincided with a strong adverse impact on crash rates in both the Midland area and the rest of New Zealand.
Another analyst performed a time series analysis of the data to take out the seasonal variation:
Again, this analysis (see Trend graph segments) highlights exactly the same three features of the data.
The LTSA response to this work which I will not publish because of privacy considerations for the analysts amounted to ad hominem attacks on the people coupled with an incompetent inability to even comprehend their work. Readers can get the flavour of it by reading the LTSA response to my original critique here: LTSA Response.