All right.
Since then, we've done a lot of research, especially in terms of the data analysis you mentioned, and we found that it wasn't a very common practice in Canada. We had to broaden our research to look at other countries. We've developed a three-step data analysis system.
In the first step of our process, we're going to apply some statistical formulas to our data. Right now, to date, we've closed about 700 contracts on our e-bidding system, worth about $200 million, and that's for our construction contracts worth less than $10 million. We were able to apply the statistical analysis to all that data.
It starts with a coefficient of variance, if you'll allow me to be a bit technical. The coefficient of variance measures the distance between bids. Research has shown that when there is a collusive environment, the bids tend to be closer together. The coefficient of variance will measure that, and if the coefficient of variance is lower than roughly 0.04 to 0.06, it is an indicator that there's a potential collusion.
Then we apply another series of data analyses, and if there's a trend showing that contractors show up on each of the analyses, then we know we need to proceed to a further, deeper investigation.
We look at the relative distance between bids. This is another measure, and research has shown that when there is a collusive environment, usually there's a larger distance between the low bidder and the second bidder versus the distance between the second and the following bidders. When the relative distance is larger than one, it's another indicator of a possible collusive environment.
These multiple supporting data analyses will then identify if there's a trend with a particular contractor or contractors.
That's generally how our system will work. Once we've done that, then we'll move to a second step, which will be the drilling down of the data to the regional, sub-regional, and local levels. Many of our contracts are at the local level, so then we'll be able to further identify the local problem or the specific problem.
Further to that, we can do a further investigation into the files. We'll look for things like whether a competitor was hired as a subcontractor by the winning bidder. We'll look at who the plan takers were. Are the ones who were bidding also taking the plans for the bidding process? If not, perhaps they got it from their colluder. That's a general description or overview of the system we've developed so far.
We are now in the consultation process with our peers and colleagues at PSPC and the Competition Bureau. We want to make sure that once we turn the light on this process, it will provide us with data that's reliable and that an authority like the Competition Bureau or the RCMP will be able to act upon.