I don't have that particular study at hand, but I can indicate that access to EI is something that's monitored regularly by the department. Every year the Employment Insurance Commission issues their monitoring and assessment report, which deals with the impacts and effectiveness of EI overall--both the income benefits and the part II employment benefits--and there is an annual analysis of the access issue in there.
There are a number of different measures that commentators and researchers use to arrive at conclusions on access. One is indeed the measure that you referred to in one of your comments, essentially the beneficiaries-to-unemployed ratio, or how many EI beneficiaries there are relative to the total population of unemployed. That is a measure, but not necessarily the most appropriate measure in terms of the effectiveness of the program, because there are significant parts of the unemployed population that, by design, are not eligible.
For example, only 70% of the unemployed in 2007 actually paid EI premiums. Almost by definition, those people are not eligible to receive benefits. Only 54% of the unemployed actually had a job separation that met the requirements of EI in that they were laid off or they quit with just cause. So there's a significant portion of the unemployed who, unless there were major design changes to those fundamental parameters, would not be eligible for benefits. The number the minister quoted refers to 82% of those for whom the program was designed to serve being eligible to receive benefits.