How do you report on numbers that are unreported? Stockwell Day says while crime rates are down, the rate of unreported crime is up, which is something that is actually very difficult to gauge. How do you find numbers on things that, by definition, do not actually have numbers attached to them? Evidently, he knows that these crimes are more common, but how does he know? Is he committing them?
Statistics can be manipulated to say whatever you want at the best of times. If you use the right graph, a commentary that tells the story you want, and a clever use of the overall numbers, you can use almost any numbers to say whatever you want. It's often up to the reader to parse what the numbers are actually saying, and in what way they are relevant to the topic at hand. I know I have seen more than one graph that has been exaggerated to tell a story the numbers don't actually bear out.
At least manipulated numbers are still trackable numbers. There's still evidence to support them, and people can easily interpret the statistics themselves if they disagree with the agenda being pushed. The trouble with Day's numbers is that it's a statistic which can't possibly be accurately tracked. The reason reported crimes are used to judge crime rates is because these are the crimes which we can confirm actually happen. There are investigations, there is paperwork, there is a required amount of proof to ensure that it's not just ghosts being chased.
The reason unreported crimes aren't the basis for most statistics is because you simply can't keep good track of them. There is no proof the crime even happened, the nature of it, and of who committed it. It's a ghost, and you can't get accurate statistics on ghosts, you don't even know if they're real or not.
The worst part of the crimes that may or may not exist is that they are being used to justify spending money on more prisons. Now, there are reasons to build more prisons overall. If facilities are becoming over-crowded, or aging to the point that replacement is necessary, it is perfectly justifiable to build more. Statistics which could be used include those on the numbers in the prison population, the reported defects in prison systems, and other things which highlight the areas in which the country's penal system is coming up short.
Unreported crime, however, is a terrible metric to judge the necessity of prisons. The reason is right in the name. If a crime is unreported, then the perpetrator is not going to be going to prison. That much is just sense, you can't imprison someone without a charge, and you can't charge someone for a crime that has not been reported at all. Unreported crime could make sense for justifying an increased police presence in an area where they are suspected to occur. They could justify education and community watch programs to protect people. As impossible to judge as the numbers are, they might make sense in some contexts.
If we truly need more prisons, that need will be measured only in reported crimes, prison population relative prison capacity, and studies on actual prison problems. If unreported crimes are a concern, it's still firmly on the investigation and police work side of the agenda. Stockwell Day isn't telling the right story for what he wants to say.