Is EI Working for the Unemployed?

The Liberals have promised to monitor the impact of the federal Budget through regular “report cards” on government performance. One key issue to look at is the adequacy (meaning inadequacy) of the EI program at a time of rapidly rising unemployment. The Budget did move very modestly on this front, adding five weeks to eligibility in each EI region, thus extending the range of potential entitlement from 19 to 50 weeks depending on the local unemployment rate.

What impact will this have? The short answer is that we really won’t know for a very long time.

In the United States, the Department of Labor release data on a monthly basis on how many unemployment insurance claims have been exhausted, and the average duration of claims. This means that there is real time tracking of the extent to which the need for benefits is being met.

Here in Canada, Statistics Canada release data monthly on the total number of regular claims and the average benefit in dollars (with some disaggregation.) However, we get no regular data on the number or proportion of exhaustees, or the average duration of benefits (except, sometimes, with a very long lag in the annual EI Monitoring and Assessment Report.)
Data on EI exhaustees and average benefit duration is readily available from HRSDCs administrative files and should be released to Statistics Canada on a monthly basis for publication, with an appropriate breakdown by age, gender, province and EI region. We simply do not know today how many workers hit by the early stages of this downturn in the manufacturing and forest sectors have already or will soon be exhausting benefits, and how many newly laid off workers are managing to find new jobs.
Going one step further, the EI administrative data could be used to track changes in wages of workers following a layoff, and to track other important labour market dynamics.
(With thanks to Laurell Ritchie of the CAW for bringign this issue to my attention.);


  • Actually, the US publishes weekly numbers on initial and continuing UI claims.

    I’ve been thinking about this as well. StatsCan has historically been more concerned about keeping revision errors small than with the timeliness of their data. Those aren’t my priorities just now.

  • Add to that the number of folks who are not elegable because they were self employed, are just entering or reentering the workforce or otherwise excluded and it can be seen that the numbers could realy be skewed.
    Recently laid off workers applying for EI may wish to view some important information in a series of posts at

  • Like I said before, I worked on that data file for quite a long time, and yes those attributes of the data that are not included in the current regime of data releases could be dug out. It is not a matter of HRSDC releasing the data, it is a matter of Statcan actually doing something with the data and putting some money into it. They came very close a few times to giving up the file and have cut the program down to be a shell of its former self.

    A lot could be done to elucidate the effectiveness (ineffectiveness) of the current EI policy from the data, however, that would take political will to ensure that such tasks are completed, and currently at both HRSDC and Statcan it is not happening. Again I go to my anti-worker bias that has been brought in by the Tories at both barns.

    It would not take much to expand that program, a couple hundred thousand dollars, maybe a little more. Given the dollar amounts that are given out, one would think that any rational institution that had some competence in delivering what the Statistics Act lays out as a mandate would seemingly be expanding EI statistics rather than cutting them.

    Both HRDSC and Statcan should be held held accountable for this malais that has come across the data

    I will also throw in Employee injury statistics into the mix of higher level complaints. The Association of Worker Compensation Boards has opened up a sweet heart deal with Statcan whereby injury Statistics are no longer monitored by Statcan, and instead are now monitored and collected by the wolf themselves. Hmmmm… have you looked at injury statistics lately, a mighty consistent drop across occupations since the AWCB have taken over the stats on workplace injuries. Marrying that up with the death dollars they have been giving out to employers and I see another huge question mark in the databases we all seem to take for granted.

    I got to cut back on my salt! No more crackers please, just give me the soup.

  • Good points.

    The annual EI Monitoring and Assessment reports also don’t report on the industry or occupation of those applying for or receiving benefits, as far as I can tell.

    I don’t see why this information couldn’t be easily collected and distributed

  • Hi All,

    I also worked on this file a while back. The NAICS code is on the file somewhere. There is also a new linkage that was achieved in the business register at Statcan that allows one to link the EI claimant with the NAICS code through the payroll number. I worked on that algorithm and it is a fairly good linkage.

    It even takes care of the head office problem when it comes to BN matching at the enterprise and establishment level, i.e. you can get a fairly accurate linkage as to the location of the Claimant and what establishment in a mutli-establishment enterprise the claimant was attached to.

    There is a lot of potential with that file, it just need to be mined as Salty above mentioned.


  • Stats are stats and the’re all BS. Take the EI area I live in Muskoka-Kawarthas. The Important stat here is Participatioin rate. If that is low, and it’s the lowest in the table, then the EI stats are all FUd. We have to track the discourages workers!!

Leave a Reply

Your email address will not be published. Required fields are marked *