Thursday, February 08, 2007

The sticky mess we’re in?

The quality of Canada’s new and existing jobs is slipping, concludes CIBC economist Benjamin Tal after calculating the latest Job Quality Index (JQI), which is based on compensation (70% of the index) and job stability. Tal suggests that the JQI is slipping because self-employment is rising (not necessarily a bad thing if it’s encouraging workers into the labour force who had previously outsiders) and/or unskilled workers are being used as substitutes for skilled workers because Canadians lack the right skills.

Perhaps an interesting statistic would be the relative "stickiness" of the JQI over time. To the extent that we can expect bad jobs to replace good jobs more quickly than vice versa, stickiness may be somewhat predictable, depending whether the index is rising or falling. But if the JQI quickly rebounds after falling, and if we know which components caused the JQI to fall, this may be telling.
For example, consider scenario A and B in period of time (t), both of which take the JQI to the same level below JQI (t-1). In scenario A there’s weak compensation due in large part to an increase in self employment spurred by individuals inside and outside of the labour market. There’s also weak stability because an increase in jobs in western Canada draws workers (again, from within and outside of the labour market) west to try on jobs until they find a satisfactory match.
In scenario B the situation is similar, only now the greatest weight pulling down the JQI is unskilled workers being used as substitutes for skilled workers when in actuality they are not substitutes. Again, consider that JQI at B is at par with JQI at A.

Would either scenario have a stickier JQI? The significance has to do with our productivity and our ability to rebound from a recession. In scenario A we have a weakish economy but more people in the labour market. In scenario B we have a weakish economy and evidence of more of a mis-matched, unskilled labour market. If we can agree that we're seeing a mix of scenario A and B, perhaps the [edit: greater ] A is, the less sticky we can expect the JQI to be. To be clearer, studies have shown, for example, that "a lack of employment stability, job skills, and occupation-specific experience can impede welfare recipients’ abilities to obtain “good jobs” or transition into them from bad ones."

On a side note, overall I like CIBC's’s JQI as a raw estimate. It has its weaknesses (no pun intended); for example, it's based only on compensation and job stability. However, it also avoids mistakes made by others (eg, CIBC uses data from 100 separate industries, rather than data from industry aggregates).


amphimacer said...

Predictable weaknesses aside (we should not expect an artifact like this to be perfect), I'm curious and (as always) concerned about how it will be used: are we going to say we need to do X or Y in order to improve the JQI? If so, is that going to improve people's lives?

When I hear economists who work for institutions like banks espousing some new regulation, deregulation, or other response to things like this index, their concern tends to be for the economy as a whole. There are some who are worried about the wide gap between the rich and poor, but most maintain that if the economy gets better as a whole, then, as the saying goes, all boats rise with the tide. Some evidence -- not all, but some -- suggests that this is not always the case. So a promising signal like this one, which may tend to speak for the people at the financial bottom end as much as for those at the top, should not be squandered. I look forward to seeing how it is brought into economic equations, and what recommendations are derived from it.

true dough said...

Thanks, Amphimacer.

I think the JQI, like the ILO’s employment slack and other such labour measures, can be used as a complement to employment rate figures. For example, if we know that the jobs being created are, say, part time, or self employed, or temporary, this tells us much more than simply knowing that employment has increased by x per cent. Not all jobs are made equal.

To answer your question, if doing X and Y improves the labour picture above what it would be in, say, a recession (eg. let X = lower barriers to self employment), than, yes, this could improve peoples’ lives. Actually, one of my theories is that the reason why self employment rates are so pronounced in periods of poor economic growth is that individuals cannot find other satisfactory work, so they use self employment as a stepping stone. I’ve found that Canada has followed this pattern since at least the early ‘80’s. If such patterns are encompassed in the JQI, than it’s a convenient tool.

Anyway, I think you’re right to be cautious about the JQI. Everyone measures it differently, especially the stability component. On a related note, I heard a man give a lecture last week and he was touting his job security index. Of course, this concept is widely used, but his index included a nations’ propensity to unionize. This means, all else being equal, that Canada’s job security would be perceived as higher than the US’s. But what happens when our high unionization falls over time? Our job security index falls, all else being equal. Out of context it sounds ghastly, and inaccurately so.

My proposition is that we should also measure the “stickiness” of the JQI, for lack of a better word. Not only do jobs differ in quality, but qualities of jobs (eg self employed, permanent, etc. jobs) submerge, re-emerge, surge and fall at different speeds and face different frictions. Suppose that A,B, and C are types of jobs with different frictions and 6(ABC) is the “natural” level of JQI. Suppose also that a JQI level of 6A + 1B + 1C might snap back to 6(ABC) faster than a JQI of 1A + 1B + 6C ever does. I’m not trying to use stickiness as an ugly second dimension to the JQI. I perceive stickiness as a weight that could be useful in other applications rather than as a simple prediction of when the former level of JQI will be reached again.