Just on the different stats kind of thing, this is a complicated thing too. But just simply, the numbers we're talking about, in terms of coverage, are very simple. It's the number of people receiving regular employment insurance benefits divided by the number of unemployed.
What the government does is this. Its counter-argument is that the denominator, which is the number of unemployed, is too broad. For example, you're including self-employed. Well, the self-employed aren't part of the program. Or you're including somebody who is new to the labour force, and they're not included.
So basically what happens is that StatsCan and the EI Commission come up with a definition with a narrower denominator. They're saying, well, not everybody who's unemployed deserves benefits, so we're going to throw them up. So that makes a difference.
What we're saying is that we believe people who are unemployed should receive unemployment insurance. It's not certain kinds of people who are unemployed; it's everybody who's unemployed.
And there are pros and cons to all these definitions. As Barb was saying, there are nine definitions of what unemployed is. But I think it is important to take a broader look at things.
Just one quick thing on what you get, compared to what you put in, in terms of premiums. That's not the way it works. For Canada and Quebec pension plans, your benefits are related to your contributions. With EI, the amount of benefits is based upon your insurable earnings, which is 55%, which is extremely low internationally. So it's not really how much premium you put in. We all put in, actually, the same premium. It's just that the benefits you get are based upon your insurable earnings.
One of the things that happened—I didn't have to chance to mention it—was that the government froze the level of maximum insurable earnings for about ten years. That meant that the maximum benefits you could get declined every year, in terms of the amount of inflation. And that's why benefits, all benefits, are lower than they were before, because the maximum is lower. It's what I call “policy by stealth”. I mean, people don't understand, but when you don't index a benefit, then its value declines over time.
Again, another problem with a program that is so complicated—