My fear: Michael Ignatieff is to EI as Stephen Harper is to the GST

Well. It would appear that I have become the go-to person for those who are interested in examining the merits of the policy proposals that are the backdrop for Ottawa's psychodrame de jour. Put briefly, my point - expressed here and especially here is that Michael Ignatieff's proposals for EI reform ignore the scholarly literature on the subject, and are more likely to do harm than good.

The Liberal talking points are the following. From the comments here, we have

"[A]pproximately no-one who has done research in this area thinks that reducing the eligibility requirements is a good idea."

Yeah, except the TD Bank, the CCPA and the Government of Ontario, all of whom have suggested flatten the hours of eligibility required to qualify for EI.

Also, the cost of a universal 360 hour standard has been pegged at somewhere between $500 million and $1.5 billion--even though it would add no additional cost for those already collecting benefits. Based on that alone it seems as though this must benefit some people. Right off the bat it occurs to me that those who might benefit most would be (a) people who just returned from parental leave before being laid off, (b) those who work seasonal jobs who will now be unable to find out of season work and (c) those who have picked up some contract work since entering the workforce but are now unable to find more. All of those people are, quite obviously, affected by the recession.

And even if reducing the hours needed to qualify for EI didn't help anyone, how would that make it a "dumb" idea as opposed to a simply a not very important issue either way? If it doesn't affect many people how can't be that harmful?

Did you consider any of this before writing your post?

Um, yes. Yes I did. To the first point, I would reply that when it comes to scholarly credibility, the collective expertise of David Green, Craig Riddell, Thomas Lemieux and Bentley McLeod (plus others whose work has passed the crucible of peer review) greatly outweighs those other sources.

The other points ignore the possibility that there might be other, better ways of spending that money. Why not beef up the GST credit or the Working Income Tax Benefit?

Elsewhere, one of my colleagues in academia passed my blog post to a Liberal party member of his acquaintance, and got this response:

1) We're not suggesting eliminating the waiting period for eligibility. We're ensuring that for the duration of this recession, the amount of hours required to qualify for EI does not vary depending on where you live in the country.  We're also reducing the minimum number of hours from 420 hrs to 360 hrs because getting EI payments (as small as they are) in the pockets of Canadians who have lost their job through no fault of their own is one of the best ways of injecting stimulus into the economy.

Again, this is the part that ignores the question of whether or not there are better ways of distributing that money.

2)  On the blog post....the problem we are trying solve is regional discrepancies in eligibility, which may make sense in an era of prosperity, but not in a recession. Because right now it is no easier to find a job in Calgary than it is in Windsor.

This is a spectacular non sequitur. If this were the concern, we'd be talking about increasing the length of the period in which people were eligible for benefits, not reducing the number of hours worked in order to become eligible.

The blogger is also wrong that nobody has done research on this.  For example, only a few weeks ago TD Economics put out a report confirming that the regional variation does not make sense right now.

Once again, I am compelled to remind readers that a TD Economics newsletter does not carry the same intellectual heft as the accumulated wisdom of leading scholars in the field. And a casual glance at the most recent Labour Force Survey suggests that labour market conditions are not uniform across the country.

And that only 60 percent of individuals who pay into the EI system are receiving benefits when they lose their job..again, through no fault of their own. You cannot get EI if you quit.  You cannot get EI if you were fired for cause.

Okay, let's think a bit about a system in which you get EI benefits if you are fired for cause or if you quit.

(Pause, while we think about it and perhaps look up the Wikipedia entry on moral hazard.)

Exactly.

It is at this point our thoughts turn to Stephen Harper and the GST. As a trained economist, Harper should have known that cutting the GST was among the worst policy proposals he could come up with. But he set aside that knowledge - and presumably the advice of people who were experts in the field - and did so anyway in order to win votes.

And now we have Michael Ignatieff, whose reputation as an intellectual is one of the Liberals' strongest points. But it's hard to see much daylight between the Prime Minister and the Leader of the Opposition these days. When it comes to choosing between the recommendations of scholars and the recommendations of political spin doctors, Michael Ignatieff looks as though he is comfortable making the same choices as Stephen Harper.

Comments

[Vladimir]

Shouldn't we be debating how to reform the method of financing EI? Perhaps a counter-cyclical payroll tax, i.e. one that would decline as unemployment increased thus providing automatic fiscal stimulus.

[Neil]

I think you missed the point on the last one. I think the writer of that comment was trying to say that there's 40% of people who have paid into EI, lost their jobs through no fault of their own, and cannot collect due to eligibility requirements. The author wasn't trying to say that people should be able to collect if they quit or are fire for cause, just making the point that they can't, so that 40% are essentially getting screwed.

(Note: I have done no research into the validity of the 40% claim, I'm just saying that your response to the comment doesn't actually make any sense.)

[Patrick]

I think Stephens point is that there are plenty of good ideas we could be debating, but that's not what the parties are doing.

Apart from the stupidity of it, I find it infuriating because there are real consequences for real people. The shrinking auto sector and collapsed manufacturing sector of southern ON is going to leave a big hole in the economy. What can be done to fill it? That isn't clear - at least not to me anyway. Unless we want to create a situation where the only growth industry is the dole, then we better start applying some serious brain power to the problem.

[Nick Rowe]

Help me think through this one, Stephen.

There are 3 different instruments the government can choose with the EI system: number of weeks worked before eligible; number of weeks of EI benefits if eligible; and size of benefits per week.

Moving each instrument, while holding the other 2 fixed, traces out a trade-off curve between the moral hazard problem and helping reduce inequality. Trace out all the possible trade-offs for each of the 3 instruments holding the other 2 instruments fixed at all possible combinations. Then draw the envelope of all those trade-off curves, to get the optimal trade-off curve.

We want to be somewhere on that optimal envelope trade-off curve. And the argument of the economists you cite boils down to saying that we are currently off that envelope, and that shortening the number of weeks before eligible moves us even further away from that envelope. By increasing the number of weeks before eligible, and loosening one or both of the other instruments, we could reduce moral hazard and improve insurance at the same time? (or improve one without worsening the other).

Or, there's a fourth instrument, GST credit or Working Income Tax benefit, which would work even better to get us to an even wider envelope.

[Chris S]

I'd think more highly of the moral hazard point if the insured had some choice in the matter. And in fact, we used to have a system where you could get EI (UI, then) benefits if you quit, and although it was different, the country didn't exactly fall over. The main (political) problem is that it cost more.

What we have now is a system that imposes an extremely high transaction cost on people moving to jobs that might be a better fit, even to the point of improving the GDP. And that is something both the Prime Minister and the leader of the Opposition should also be encouraging.

Instead, people hang on for dear life to jobs they don't like, even when they think there is a good chance they could be doing better elsewhere. Moreover, since the cap on insurability is lower than many wages (~39,000) and the limit on payouts (even when you do get coverage) is so low (~55%), you will *still* suffer notable negative consequences.

We can suggest a better system might allow people to choose whether or not to use EI, but of course that would muck up a major stream of payments to whatever government is currently in power. And, of course, we can see problems looming there - people with the lowest of assets and wages ending up with no insurance because they decided that food and shelter were more important than paying premiums.

If you're going to enter into this debate, I think you need to be prepared to defend your suggestions on two other bases; one, the political acceptability to at least one major party, and, two, be prepared to discuss significant interactions with other aspects of our existence, economic and otherwise.

I don't think it's sufficient to intelligently criticize the government's actions. You need to be able to discuss palatable alternatives. And in that sense, lowering the elegibility works. It is relatively low cost (compared to extending benefits, or raising the wage-related rate) and it is seen as being a strong net benefit to people when they think of the consequences of losing their job. Moreover, it's one change that can be implemented quickly; there's a low transaction cost to the government to the change itself - both monetary and political - compared to many other possible changes.

Your other suggestions (the GST credit and the Working Income Tax Benefit) both end up typically coming at the end of a tax cycle - in other words, if the government made the change tomorrow, and someone is out of work the next day, it will likely be 10 months (next April) before they see any money related to those changes (or related to those programs at all). Although they might make lots of sense in a somewhat abstract policy sense, they aren't going to work if there is no political will, or if the individual's net year-by-year benefit corresponds to a month-by-month problem.

[Nick Rowe]

Neil and Chris: if there really were no moral hazard problem with shortening the number of weeks before eligibility, then how do you explain the spike in weeks worked that exactly matches the number of weeks before eligibility? Is it just a fluke? And a double fluke, if, as in the paper cited by Stephen in his first linked post, that spike just happens to move exactly the same number of weeks as an exogenous change in the number of weeks before eligibility?

It might look like there shouldn't be a moral hazard problem, but the data tell us there is.

Here's how to resolve the paradox. I'm an employer who wants to attract the best workers I can at the lowest wage. If you need 10 weeks employment for eligibility, I offer jobs that last 10 weeks, and tell workers they will be laid off after 10 weeks. At the end of 10 weeks, the workers are laid off. They have no choice in the matter. They can't help getting laid off. They didn't quit, or get fired for cause. But I the employer chose to offer a 10 week job because I knew that some workers would find a 10 week job very attractive, because of EI eligibility.

You want hamburgers? I'll make you hamburgers. You want 10 week jobs? I'll make you 10 week jobs.

[Nick Rowe]

Sorry Neil. My above point should have been directed just at Chris (and at the Liberal party member Stephen cites). You were making a different point.

[Mark Dowling]

Chris S - opting out of EI need not necessarily mean no EI.
1. Ontario gets its workers out of EI into OntarioEI. Pricey to set up but Quebec has QPP so it can be done, and Ontario's workers are no longer paying into a fund on an equal basis but getting benefits on a differential basis.
2. Employers are allowed to pay the premiums to a bonded private insurer who provides EI on a contracted basis. There would be a portability issue here however.

In neither case above would an direct employee be allowed to simply dispense with EI. However, self-employed people - doctors, for instance - can do so at present and government seems to find that tenable.

@Nick - would it be possible to claim that the employers, by setting up 10 week jobs, are not actually firing their workers, as in the recent case involving a staffer to Sen. Colin Kenny?
http://ago.mobile.globeandmail.com/generated/archive/RTGAM/html/20090506/wretire06art2218.html

[Nick Rowe]

Mark: Neat link! Yes, if the employer deliberately sets up a 10 week job, because he knows that's what the workers want, then they really quit, rather than being laid off. But you could never prove it in any individual case. Only when you look at the spike in the data can you prove it in aggregate. So they are eligible for EI, because you can't prove they didn't want to work more.

[Patrick]

In my view inequality is important but it seems to me that the main economic argument for EI is to provide workers with bridge financing to the next job that best uses their skills. Without EI, many liquidity constrained households (which I suspect is the typical household) would quickly default and be forced into insolvency and/or workers are forced to put their skills to sub-optimal use (e.g. PhD's driving taxis). This would tend to make recessions worse wouldn't it?

Scott Sumner, in a previous thread, mentioned Singapore using forced savings for this sort of thing and it got me thinking: why not have a system where people who find themselves unemployed can withdraw from their RRSP without taking a tax hit, credit what they withdraw back to their allowed contribution, and then have a system that automatically repays what they took out, with all the normal tax benefits of a normal RRSP contribution, once the person starts working again? Then the government could say; first you pay from your savings up to some amount - and BTW we give you a pretty good tax break on repaying yourself. After you hit the limit on using your savings, then EI kicks in. While we're at it, a similar system for financing a new business might also be good. CRA has the Home Buyers Plan and the Life Long Learners plan whereby you can dip into your RRSP for buying your first house and financing education, and they work quite well in my experience (paid for my wife's master's degree that way!). So why not do it for dealing softening the blow from unemployment and starting a new business? Those are the other two 'big ticket' items people save for.

[Stephen Gordon]

That sort of thing might be something to think about for self-employed workers. Self-employed workers aren't eligible for EI, presumably because of the moral hazard problem: it's hard for an outsider to monitor effort. But it's not entirely satisfactory to leave them all to their devices.

[Chris S]

I'm probably guilty of using excessive wording, but - I expressly did not dispute the existence of moral hazard. I simply don't think that it's such a big deal. You could try being the type of employee who regularly uses such a program to hop from job to job, but that's going to be on your job record (or you don't apply using a job record) and you're going to get jobs accordingly. You might abuse it in the short term, but that's about it.

I didn't go in to it, but I strongly suspect that right now, employers often simply fire people *not* for cause even when they could, because they know that when fired for cause, employees push back because their benefits are on the line. It takes time and effort to document for-cause, and many employers would likely rather just wipe their hands of it, and let the ex-employee have the benefits. In that sense, the current program is already subject to moral hazard in a sense - employers benefit from an easy ride in immediate firings, when they shouldn't.

I see things like this on a regular basis. I find it interesting that there is a common set-point for a pair of eyeglasses is $200, and the usual minimum vision care amount in an employer's health insurance package is $200. Wow, how fortunate that the market price of a pair of glasses all over the place just happens to be the same, and be the same as the insurance, too! (Bridges available on request...) Somewhere out there are eyeglass vendors benefiting from my insurance when they shouldn't!

I think Patrick is giving an example of where I would rather be headed. We could leave every facet of the program the same, except for allowing it in the event of voluntary departure, and call it the Job-2-Job Bridge program. It's still limited, but it's more a shared risk pool, with the expectation that in the interests of a well-functioning economy, the payouts may benefit people based on their own actions. It gives you the chance to step away from your current job, get some appropriate retraining, and take up a well-planned job search when it makes sense - like, when employers aren't so risk-averse that they wouldn't hire people for free.

It's a big jump in optics, but if the cost increase (yes, there would be one) is not too high, longer term it will be at least partly balanced by an improved GDP and the taxes that generates. (We'd never prove it, I agree.) A program with a chance for broad public acceptance, only a moderate increase in funding, and a strong rationale for improving the overall functioning of the economy?

Buried in all of this I find at least one common theme - the complete step around transaction costs seems to crop up again and again in economic analyses. It costs something to find a job, leave a job, start a job, change jobs - so much so that even if two people would be better suited to switching jobs, it's likely not going to happen because of the transaction costs. At the aggregate level, that's a problem we should be trying to fix - but instead we're ingraining it. (Parking meters and road tolls are often truly egregious examples of ignoring transaction costs.)

Patrick, the RRSP example of bridging *almost* works - except for people on defined benefit pension plans. Those people don't have RRSPs - nor do they have RRSP contribution room, because there is a special deduction to account for the defined benefit pension. They have the pension, but not in a form they can draw down against. How would you accomodate those people?

[Nick Rowe]

Your eyeglass example is great Chris. How come, when we look across the frequency spectrum of eyeglass prices, we see a spike at $200, just like the spike at 10 weeks!

Your second paragraph, about employers finding it easier to "lay off" when it's really "fired for cause" sounds plausible. Can't see how to test it though. Maybe see how the ratio of layoffs vs (fired + quits) varies across weeks worked, or something similar?

[Phillip Huggan]

Harper's 2% GST cut costs $12B annual in revenue and stimulates production of many luxury goods that are being abandoned in the USA now. This Iggy plan costs $1B/yr, no comparison.

Raise money with cap-and-trade (give provinces jurisdiction to go to carbon tax?) revenues; instead of making it revenue neutral put revenues to killing deficit. Semiconductors companies make microchips, the product with the longest supply chain. So favour these companies for highest GDP pop (much of it to Tiger economies). If debt forces the USA DoD to scale back aerospace orders, shift to bolstering domestic space agency programmes (research labs get the dollars given now to DoD suppliers assuming scale-back). How about a national wind turbine grid (employment intensive supply chain much of it in Denmark)? Have every point in the country potentially servicable by wind turbines, I think the powerline loss limit is 1500km. Will someone please give Zenn a sliver of the public dollars that have sustained the oil sands industry?

[Patrick]

I doubt the 'lay-off' that is really 'fired for cause' is a big deal. Keep in mind that 'fired for cause' is really serious stuff. You have to be truly dedicated to get fired for cause. It's not the same as 'doesn't really do a great job' or 'I don't like him/her'. I suspect giving employers and employees a mechanism to make a graceful exit probably increases liquidity in the labour market and cuts down on lawsuits.