What is the problem that EI reform is supposed to solve?

If asked, people might say something along the lines of this Toronto Star editorial:

The rolls of the unemployed in Canada continue to grow... Many or most of them won't be eligible for employment insurance (EI), due to the program's Byzantine eligibility requirements.

Very alarming - especially that word "most". But how do we know that a large fraction - let alone "most" - of the newly-unemployed are in fact not eligible for EI? Isn't that something we'd like to know?

I know it's something I'd like to know, so I decided to do some digging. It turns out that the EI system is holding up pretty well - in fact, much better than one might have expected.

Firstly, let's look at how EI works. A newly-unemployed worker is eligible for EI if she

1. contributed to EI, and
2. had a valid job separation (i.e., did not quit or was not fired for cause).

Of course, as we all know by now, she could be eligible for EI but still not receive benefits if the number of hours worked doesn't meet the minimal criteria. Here's a breakdown of the number of unemployed people in 2008, according to a StatsCan survey on EI coverage released last month:

Non-contributors: 327,500
Contributors with an invalid separation: 195,300
Contributors with a valid separation but not enough hours worked: 102,100
People receiving benefits: 469,700
Total unemployed: 1,094,600

Here, we embark on a round of arithmetic. Some people might be tempted to divide the second-last number by the last, and conclude that 100*4697700/1094600 = 42.9% of the people who were unemployed in 2008 received benefits. That doesn't sound like very many, until you recall that this includes people who didn't contribute (29.9%)  and those with invalid separations (17.8%). If you want to embark on a crusade to extend EI to these groups, then you might want to take that 42.9% number seriously. If you don't - that is, if you think that EI benefits should go to those who contributed to EI and who are unemployed through no fault of their own - then that percentage is meaningless.

But if you limit attention to those who contributed to EI and who had a valid separation, the relevant coverage ratio is 100*469,700/(469,700+102,100) = 82.2%. SInce 2003, this ratio has fluctuated between 80 and 84%. (Update: see here and the links therein for a discussion on the reasoning - and the notable lack lack thereof - behind reducing the hours-worked requirement in order to increase this ratio.)

So what could we have expected of the EI system when the recession hit? Since the October 2008 peak, employment fell by 369,900 as of June. Suppose that everyone who lost their job since October 2008 satisfied these criteria:

• they had contributed to EI, and
• they had a valid separation (i.e., had been laid off, not quit or been fired for cause)

If the EI program had performed according to previous experience, we'd expect the number of EI recipients to increase by 80-84% of the loss in employment, that is, between 285,900 and 310,700. In October 2008, there were 500,340 people receiving regular EI benefits, so means that June's numbers should have been in the 786,240-811,040 range. And that's only if everyone who had lost their job had contributed to EI, and if none had quit or been fired for cause; if neither of these assumptions hold, then those numbers would be smaller.

Today, StatsCan released the number of people receiving regular benefits in June: 816,600. This number is - as far as I can tell - impossible to reconcile with a narrative in which EI is systematically failing to provide benefits to the newly-unemployed.

These are of course back-of-the-envelope calculations, and we really should look at the administrative files to get a more precise picture. But if there were first-order problems - the sort that we might reasonably expect to be the basis of an election - then they should show up in the macro data.

The EI system is outperforming what we could have reasonably expected from it. So why exactly does anyone think that it needs fixing?

Comments

[Two Hats]

A cogent analysis, but why assume that the 15-20% of contributors who routinely don't get benefits, shouldn't? In our hyperbole-loving culture, 1-in-20 (5%) 1-in-10 or 1-in-5 (20%) don't sound significant, but they are. Moreover, there's a long-standing problem with the system, so having it match (or slightly beat) historical performance isn't saying much.

Personally, I'd like to see EI split into a true, actuarially-based employment insurance and a separate support program for seasonal workers (and probably a subsidy for those for whom the new insurance premiums are onerous), but so long as we have the system we have, it's worth breaking down that 15-20% who pay for no benefit.

[Stephen Gordon]

I'd covered that point many times before, but I should have addressed it for new-ish readers. I've added an update to this post, which points to other posts that expound on this issue.

[Erin Weir]

Julian Beltrame of Canadian Press made the same point in reporting on this morning’s EI statistics.

[Stephen Gordon]

Ah. Indeed he did, and well done him.

Sadly, he then lapsed back into 'shape-of-the-earth-opinions-differ' reporting for the rest of the article...

[Erin Weir]

Something I have not completely gotten my head around is how much seasonal adjustment is increasing the reported number of EI beneficiaries. For June, seasonal adjustment added 86,000 to the number unemployed and 154,500 to the number of beneficiaries. Without seasonal adjustment, an appreciably smaller portion of unemployed Canadians are receiving EI benefits.

[Stephen Gordon]

As I said, these are back-of-the-envelope numbers. If we were talking about a OMGWTFBBQ WE MUST HAVE AN ELECTION NOW NOW NOW WE CAN'T WAIT NOW NOW WE MUST FIX THIS NOW situation, it would be pretty evident no matter what data we used.

If you have to go to seasonal adjustment factors after 8 months of recession, that sort of reduces the force of your argument.

[Erin Weir]

I was asking a genuine question, not posting in block capitals.

If your only interest is dismissing proposals to enhance EI, then keep it up.

If you are interested in providing some insight on today’s numbers, I would welcome it.

[westslope]

If Unemployment insurance and Employment Insurance programs can be extended to self-employed commercial fishermen and young mothers who become pregnant and have children--almost universally a planned event these days--as well as politically powerfully workers like CAW-organized auto workers who are "laid off" for one day a week, why not contort the program to transfer income to new categories of 'workers' who might be convinced to vote for the Liberal Party of Canada under Michael Ignatieff's leadership?

Question for Two Hats: Why do you want to subsidize seasonal workers with income transfers disguised as insurance benefits? Is your objective to decrease the social return of agriculture, forestry, oil & natural gas exploration, tourism, etc.? In most cases, that means decreasing the social return on valuable public assets. Is that really what you had in mind?

---------------------------------------------

I understand the appeal. I spent part of a summer once enjoying a magazine freelance, flyfishing and wilderness backpacking arts grant from the Canadian federal government (before returning to university) thanks to the Unemployment Insurance (UI) scheme. It was enjoyable, very refreshing in fact. That and my writing skills improved, and I picked up a few dollars selling magazine articles.

There are plenty of life-style folk, especially in the province of British Columbia, who would love to see the seasonal component of EI extended and made more generous.

[Stephen Gordon]

I was asking a genuine question, not posting in block capitals.

If your only interest is dismissing proposals to enhance EI, then keep it up.

If you are interested in providing some insight on today’s numbers, I would welcome it.

I went through the numbers, I acknowledged that they are rough approximations, and I said that if there's not a significant first-order problem there, then those who would claim that reducing the eligibility requirements have to make the case that there is a problem, and that their solution doesn't make things worse. If the issue becomes one of expanding benefits because EI recipients are exhausting their benefits, then I'll have other things to say. I'm not averse to EI reforms that address real problems.

But if your starting point is that EI should be extended to people who did not contribute to EI and to people whose separations from employment do not satisfy the EI requirements - that is, if you think that repeating the 51.3% number has any meaning - then it's hard to know just what it is you're up to.

[Stephen Gordon]

Heck, I'd like an answer to the question that forms the post's title, come to that.

[Erin Weir]

I was just trying to prompt some analysis beyond what Canadian Press posted earlier today.

As you know, I support lowering the number of work hours needed to qualify for benefits. That has nothing to do with people who did not contribute, quit voluntarily, or were fired with cause.

The 49% of unemployed Canadians currently not receiving benefits does not only consist of people who did not qualify based on criteria that you deem appropriate. It also includes people who initially did qualify but have since exhausted their benefits. In other words, extending the duration of benefits (which we both support) would increase the 51% figure.

It still seems to me that, if your goal is to help the unemployed, you might consider putting your energy into proposing a longer duration of benefits rather than into opposing any increase in the accessibility of benefits.

[Stephen Gordon]

Why? It's possible to hold both positions. If I'm more vocal about the accessibility issue, then the reason is that it's the one that people keep talking about. It's a bad policy; the only way to persuade me to stop denouncing it is to take it off the table.

Is there anyone *opposing* extending durations if necessary? There's already been an extension, and I don't get the impression that there's been a line drawn in the sand on this issue. If it happens, I'll get on that as well. But until then, I don't see the point in mounting an attack on a position that no-one is defending.

[Patrick]

At the risk of stating the obvious: it's about trying to find an issue that has traction with voters. Add it to the list of rubbish we're sure to hear about from all sides.

Ugh.

[David]

Maybe this is a little off-topic, but as a university student who has worked full-time jobs each summer, I get frustrated paying EI premiums even though in all likelihood I'll never claim EI because I don't aquire the 600 hours worked until near the end of a four-month summer of full-time work, and because when I leave to go back to school I quit my job. So if we're talking about reforms, maybe there should be an opt-out clause for full-time students on summer jobs, or have a way to claim it back on tax returns. It seems unfair to be forced to pay into an insurance program when I'm not eligbile for any benefits.

[Erin Weir]

In theory, I think it probably makes sense to give students their EI premiums back. For every two months enrolled in a full-time post-secondary program, someone could be refunded one preceding month of EI premiums.

However, EI premiums are only 1.73% of earnings (and are already tax-deductible). The minimal value of student refunds would have to be weighed against the cost of administering them. Also, employers would undoubtedly use this precedent to try to get out of paying their EI premiums (2.42%) when employing summer students.

[Matthew]

Fair enough Stephen, but I am assuming the 82.2 percent figure is a national one. I suspect it would be considerably lower in areas where the hours worked threashold is higher (i.e. Calgary, Vancouver). So while nationally things look pretty good, I can understand the western premiers getting a bit cranky when people are added to the provincial welfare rolls becuase they havent worked enough hours to satisfy the requirement.

[Erin Weir]

EI benefit coverage is now lower in Ontario than in any of the western provinces.

[Two Hats]

DR Gordon: I take your point. I suppose I'm approching this from the POV that the EI system is messed up and needs repair, and that this has nothing to do with the current recession. I don't necessarily agree dropping the hours-worked requirement is necessary. I do think having variable requirements across the country appears grossly unfair and requires serious justification, which I haven't seen (nor looked for, in fairness). I suspect there are people paying into the system who can never benefit, and that's unfair. Equally unfair, there are people (eg seasonal workers) who pay too little in for the benefit they receive, because their situation is fundamentally incompatible with the concept of insurance.

I don't have an answer to your title question -- in fact, I'm on the whole against both the opposition and government proposals to modify EI; I don't think any address the problem.

I definitely agree w/ Dr Rowe in the comments you link to that tax rates (and legislation generally) should avoid step functions.

Westslope: I'm specifically proposing that these benefits *not* be disguised as insurance -- that is, realising the political reality that no one is going to be cutting off seasonal workers' benefits any time soon, we'd best at least call them income support, and not EI! And the only reason I propose *that* is because of the aforementioned political reality.

[Just Janice]

I think the biggest problem of EI is the adequacy of benefits in light of wages and likely duration of unemployment. EI reform is needed because at present benefit levels, many are likely to struggle severely if they do face unemployment. 55% of a maximum of $42,300 and a maximum weekly benefit of $455 (before tax) means that a person earning the maximum or more might have a hard time keeping a roof over their head, food on the table, and a car on the road. (Let's assume a car costs $450 per month for payment, gas, insurance and maintenance, housing is just $800 per month, and utilities are $200 per month: there's just $521.66 per month to pay taxes and food)...if you've been used to earning $70,980 per year, can you get by on about 33% of your usual earnings? I'd likely be willing to pay more in premiums to extend the cieling on earnings to something that is more reflective of a larger segment of the population (ie. set it at 80th percentile of Canadian taxpayers...). The inadequacy is even more pronounced in larger Canadian centres where wages may be higher, but cost of living is higher yet...

There's also a portion of people who pay but cannot access benefits (ie. the students), for these people it would be nice if they could get the EI premiums they paid back upon proof of enrollment in an educational institution...

[Matthew]

Erin Weir:

huh, interesting, thanks.

[Phillip Huggan]

Kind of related and timely to thread: it might be cost-effective to federally permit an added week of paid sick-leave over the potential 2009-10 Swine Flu season. The logic being people would be less likely to infect co-workers. Depends on whether R0 (the component of it that is physiological transmission) is low enough for this strategy to reduce # of infected or more likely reduce peak load; thin out the peak of flu infection graph.
IDK if honour principle or later antibody test or whatever, would be enough. If workers can be educated about early flu symptoms, is another issue.

About EI: temp labourers pay into it and can't be fired, can't collect. Two past Cgy jobsites interested in hiring me insisted I first quit my temp employment before applying, which would've rendered me invalid for EI or welfare. The latter is against my Charter Rights though Harper and the Aspers killed my ability to stand up for my Human Rights here by killing the $27M/yr Charter challenges programme and replacing with $6M/yr private surrogate (funded from $4B in Canwest debt I assume; what a tiny Foundation given size of debt).

Then again, 2000 election called because Right just abandoned flat income tax and had Creationist candidate (like our current Science Minister).
2004 P.Martin hubris.
2006 determined by phony (illegal?) RCMP investigation.
2008 called either when media cabal fix established potential majority (platform included raping Indian youths), or fear of recession optics.
An election over EI "reform" would be a contemporary Canadian record measured by relevancy.

[Alex Plante]

When I graduated with a Civil engineering degree in January 1993 there where 6 advertized civil engineering jobs for 300 graduates in the province of Quebec.

Because I had taken a co-op program, I was eligeable for EI (or UI as it was then called). As I had still not found a job by the spring I had a crazy idea to drive to Vancouver in my 10-yr old car, find somewhere to camp just outside the city, and try to find work there, as it seemed to be the only place in Canada where construction had not ground to a halt.

I was surprised to find out that my UI would have been cut off for the duration of the trip, and that to continue receiving it I would have to establish a permanent residence in BC.

I don't know if things have changed, but I remember thinking how stupid the UI program was, they should do everything to encourage people to move to where the jobs are.

[Matthew]

"I don't know if things have changed, but I remember thinking how stupid the UI program was, they should do everything to encourage people to move to where the jobs are."

I think that this alone is a pretty good argument for EI reform. Yes, the current system reflects the fact that there are different economic circumstances in different areas of the country, but they also interfere with the efficient functioning of labour markets - further encouraging people to stay unemployed in area A while there may be a dearth of labour in area B.

[tyronen]

IIRC, EI was once available to those who left their jobs voluntarily or were fired. This was abolished by the Mulroney government in 1990 or so.

The intent was presumably to prevent people from "abusing" the system by deliberately vacationing on EI. Unfortunately, another effect was that workers were left less of a way out in cases like sexual harassment, unsafe working conditions, whistle blowing etc. Also, many low-wage employers withhold the last paycheck or two to a departed worker. Put the two together and you end up with a very intimidated, docile workforce. Maybe that was the point.

It's interesting how policymakers always try to make the labour force "flexible" - for employers only.

[Patrick]

More info on EI qualifying here:

http://www.servicecanada.gc.ca/eng/ei/information/voluntarily_leaving.shtml

You can quit your job for sexual harassment, unsafe conditions etc and still qualify. BUT you need to take 'reasonable' steps to deal with the problem before you quit. For example, if someone is harassed, EI would probably expect to see documentation that you went to the union or the HR dept, filed a police complaint, or even better a lawsuit. Similarly for unsafe work, they'd expect to see some evidence that you sought remedy before quitting.

If you quit your job to move to keep your family together (e.g. spouse gets job in different city), that's OK.

BTW, You are disqualified for being fired for *misconduct*, but misconduct is NOT incompetence, or not kissing-up to your boss etc... It's stuff like punching a co-worker or theft. If someone looses a job because they aren't able to do it well, despite a good faith effort, they would still qualify for EI (supposing they had enough hours). You have really work at getting fired for misconduct.

[Erin Weir]

Tyronen’s point about flexibility is an important one. If we are serious about enabling people to move around in pursuit of better, higher-productivity employment, then there is a good theoretical argument for making EI benefits available to workers who quit their jobs voluntarily.

As long as the economy is operating below full employment, the costs of abuse are minimal. If someone chooses to collect EI rather than working, that choice just leaves a job available for a more willing worker.

[Matthew]

"...there is a good theoretical argument for making EI benefits available to workers who quit their jobs voluntarily." Maybe, but don't most people only quit a job once they have secured employment elsewhere.

One idea that has been kicked around in the context of labour productivty is an interprovincial job-search tax credit.

[Two Hats]

Patrick says: "For example, if someone is harassed, EI would probably expect to see documentation that you went to the union or the HR dept, filed a police complaint, or even better a lawsuit."

I'd like to point out that small-to-medium firms typically don't have unions or arms-length HR dept to deal effectively with such a case, and police complaints are lawsuits are a *huge* step up and require a much higher bar of misconduct than hopefully exists in most harassment cases. Unless you're in a large company (large enough that you can get away from the influence of the people involved), typically you have the choice of leaving or putting up in the face of mistreatment (unless it's really beyond the pale and you can force the others to leave).

[Patrick]

That's too true. Life has a tremendous capacity for generating injustice. I don't think anyone is disputing that.

Apparently it is possible to fight and win:

http://www.ei.gc.ca/eng/board/favourable_jurisprudence/voluntarily_leaving.shtml#harassment

My view is that the bad old days of people gaming the system was sufficiently destructive and costly to warrant limiting EI to the involuntarily unemployed and then work out the exceptions.

[Phillip Huggan]

I'll parrot ad nauseam circa 2003 only half of temp/casual labour firms in the land bother to report employee earnings to then CCRA. The ranked #1 Crown programme Charter (same Charter many developing nations use as a model rather than USA's Bill of Rights) Court Challenges Programme would've given me the outlet to challenge what I thought was Human Rights discrimination.
To me people being screwed over by EI who don't have prominent last names need an outlet like the Court Challenges Programme to collect fair share of EI. I'll never know if my wit would've overcome weaselly silverspoon lawyers because Harper and Aspers killed my (and many African nation's) Human Rights here...

[westslope]

Two hats wrote: Westslope: I'm specifically proposing that these benefits *not* be disguised as insurance -- that is, realising the political reality that no one is going to be cutting off seasonal workers' benefits any time soon, we'd best at least call them income support, and not EI! And the only reason I propose *that* is because of the aforementioned political reality.

Two hats: I agree. Removing the fog of smoke 'n mirrors from bad policy is an important step in the right direction. Funny, how western intelligentsia climbed all over the Bush II regime for obfuscating policy rationale yet seems to provide conservative left-wing activists and environmentalists with a free pass to obfuscate as deemed necessary.

[Phillip Huggan]

Another EI improvement Canadians would overwhelming love but will never be considered:
If Parliament is prorogued for political and not National Security reasons, MPs go on EI instead of continuing to collect $2000/week + expenses. It is the rule everyone else has to play by.