« John Cochrane on Monetary Policy with Interest on Reserves | Main | Teaching general principles of macro »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Now, why didn't the tories, instead of making up some model that no one would ever read (and which was wholly meaningless anyhow), just publish that chart (starting in 1990) and go "ta da!". I mean, god, there's an argument that voters can sink their teeth into.

I can't help but think that it is very strange that "job creation" is the main(?) metric by which governments' economic policies are judged. Assume, reasonably, that leisure is a normal good. A policy that made us all richer would tend to reduce employment. And some very bad policies would tend to increase employment: "I know, let's throw away all the assets in people's pension funds, so they keep on working long past 65!" Lots of good and bad (economic) policies would have no obvious effect on employment, either way.

GDP per capita would be a better metric than employment (though still, obviously, subject to numerous counterexamples).

I am thinking the NDP will not think much of this chart.

I have to say I hate this kind of thing, though it is the focus of attention. Odds are the NDP was unlucky: hit by the US recession, by the desire to squeeze out inflation and an economy that gained strength as the Cdollar dropped during the early 90's. But putting that interval aside, the data shows classic small effects. It never ceases to amaze me that people spend so much effort arguing over the differences in policy when there's so little measurable difference in these outcomes - or how in the US, many people vote for the party whose results are measurably though not terribly worse for them. Put those together and you get a lot of noise about stuff that people don't process rationally.

Ideally, we would want a metric that combines a number of indicators such as employment levels and employment growth (including a differentiation between full-time versus part time employment)and of course GDP or GDP growth. If only per capita GDP growth was the metric, the ranking in the above chart would probably not be that much different as it was much lower in the early 1990s and after 2000. We may also implicitly view these types of metrics from the comfort of our own situations - seeing GDP per capita as a better metric than employment growth might not appeal to people who are currently unemployed. On the other hand, if you have a minimum wage paying job, then a better job with higher income would be an obvious step up.


Presumably the logic is that "jobs" are something tangible that voters can relate to, whereas GDP (or GDP per capita or growth in GDP per capita) is more abstract and doesn't translate easily into every day life. It's the modern version of "a chicken in every pot".

Back when I was politically active (with the Tories) I attended a campaign school where we were told to prepare political documents or messages on the assumption that your reader or viewer has a grade 9 education. Now, that sounds dismissive of voters, but the point wasn't that voters are stupid, but rather that for normal people, politics is not a top-of-mind thing. They're thinking about their jobs, or their kids, or the community play, or their begonias, or whatever (in contrast to the sort of political junkies who infest political parties for whom politics is everything). When it comes to allocating scarce mental resources and energy, reading apamphlet or listening to a radio ad is not a top priority. With that in mind, political messaging has to be a simple and tractable as possible. It's great to come up with a conclusive proof that, say, Tory governments result in statistically significant improvements on all axis of human welfare vis-a-vis the Liberals, but if no one reads it, it doesn't mean much.

No doubt you and I can come up with a better metric of economic performance than jobs, but it's hard to come up with one that's as easy to understand and interpret.

Mind you, the other consideration is that most politicians are economically illiterate, so no doubt the fact that "jobs, jobs, jos" is a message that they can understand plays an important role as well.

If you wanted to really think about the effects of the parties on jobs, it would seem to me to be better to use a statistic that measures the difference between Ontario job creation and Canadian/US/OECD job creation. Something like the "Moneyball" advanced stats that attempts to remove random effects and the effects of other players.

Of course, this is going away from the one dimensional simplicity of using a jobs chart to a party's advantage, and to engage a voter on a simple level.

On the other hand, if you are interested in how a party might effect jobs (and not GDP, well being or whatever) it would seem to be more relevant.

No wonder people have little respect and trust in economists.

Economists such as yourself should know that policy takes time to have an effect and there are much more important factors in play than who's governing when it comes to job creations. Yet, you mislead people thinking there's is a very strong correlation between governing party and job creation. Surely, you mentioned there are other factors in play near the end of your blog, but that's just like any fine print on an ad, likely to be ignored by readers and designed not to be seen.

I’ve been following the job creation debate with great interest, but unfortunately nobody has raised the main issue regarding employment growth forecast in the next decade, the impact of demographic change. Yes looking at past employment growth to evaluate future employment growth is interesting, but demographic in the coming decade is going to be very different to what we have seen in the past and this is going to have a major impact of job creation.

I’ve been forecasting long-term employment growth for the last 7 years in my daily jobs so maybe I am more aware of these issues that most people are. Let me just demonstrate my point with some simple calculation.

Based on CANSIM table 282-0001, the Ontario LFS source population was 11.2 million in 2013. Taking the growth rate for each 5-year age group from Finance Ontario demographic forecast, http://www.fin.gov.on.ca/en/economy/demographics/projections/table10.html, the source population is projected to grow to 12,207,000 in 2021, a 1 million increase. According to this projection, the share of people aged 55 and over will increase significantly; people in their 30s will also see an increase in their share, every other group a decline.

In 2013, the employment rate in Ontario was at 61.4%, employment level was 6,880,000. To really take into account the impact of demographic change, you have to use 5-year age group employment rate and population growth. If you make the assumption that the employment rate for every 5-year age group will remain unchanged from their 2013 level, apply them to the 2021 projected age structure, and you get a 59.1% employment rate, this mean 7,213,000 jobs in 2021, a 333,000 job growth from 2013.

Ok, the 2013 employment rate is probably too low. If you take the 2008 employment rates, and do the same calculation, you get a 59.7% employment rate and a 403,000 job growth. What do we need to do to reach 1 million jobs? We need to increase the employment rates of all the 25 and over age group 4.0 basis points above their all-time high. For the 15-24, I made the assumption that their employment rate will increase by 3.0 basis points above their 2013 level. This would bring the 15 and over employment rate of 64.5%.

This scenario is very unlikely. If you look at the Conference Board 2014 provincial long-term economic outlook, they have a more optimistic population growth, reaching 12,340,000 in 2021, a 65.8% participation rate and a 5.6% unemployment rate, which translate into a 775,000 job growth. Given their population and participation rate forecast, to add an extra 225,000 job to reach the one million mark, the unemployment rate would have to decline to 2.9%.

All the politics proposed by the Conservative to reach that million jobs are related to labour demand. The problem is not labour demand, but insufficient labour supply. There is simply not going to be enough working age people to create one million jobs.

Martin's number-crunching, especially with a graph, especially for Canada as a whole, would make a really good post. I really like how you have broken it down into 5-year groups, because I think that is important. Not all 15-64's, and not all 65+, are the same ages, and those groups are far too coarse for the effect of current demographics on participation rates.

But Stephen did do a related post (about 5 years back?) warning us of this demographic transition, and saying it was starting now (i.e. then).


I think you're misreading Livio' post. I don't think he's suggesting that there's a particularly strong correlation between political regime and job creations. Certainly, he doesn't say that - no reputable economist would infer a correlation (must less a causal relationship) from a three observation sample. Indeed, on the contrary, his graph suggests that there's no obvious correlation between political party and job growth since both the peterson liberals and Harris tories had more or less the same record.

Now, given that memories of the Peterson years are fading,it's entirely possible that the voters might reasonably look to the Harris years as examples of Tory economic competence vs the current incompetence of the Liberals (though god knows why the Tories haven't been trying to exploit that). That may not bear any ressemblance to reality for all the reasons Livio refers to. But so what? After all, Ontarians associate the Mike Harris tories, for example, with cuts to health care spending, when in fact the only time Ontario's health care spending was cut in the 1990s was under the Rae government - a government associated with being a collection of fiscal wastrels (which, in fairness, they were before the lenders cut off the taps) but pro-health care. But that's a political point, that perception is often more important than reality.

A comment a bit late for computer reasons
Livio: apart the fact that Bob Rae had nothing to do with John Crow creating the Great Canadian Slump and Harris/Eves reaping the inevitable upswing as a result of Gordon Thiessen restoring sanity to the BoC, what is the point?
Ok,I know,for the LIVs,government is a hazy concept and who does what is unfathomable...

Bob: grade 9? You were over estimating...Our choice in government is between a demcracy where 90% of the voters need to be retrained each morning in using a spoon for their corn flakes and an oligarchy so corrupt it would made a babylonian satrap nauseous.

This being said, it is replay of the last QC election. Promises to "create" 250 000 jobs, even though it is merely keeping pace with the labor force, talking about a jobs problem even though the employment ratio is the highest in history (but advocating the need for increased immigration...). And forgetting that no provincial government can influence short term employment in any meaningful sense...

The LIVs? My name as a descriptor. Isn't blogging fun.

JRG: "grade 9? You were over estimating...Our choice in government is between a demcracy where 90% of the voters need to be retrained each morning in using a spoon for their corn flakes and an oligarchy so corrupt it would made a babylonian satrap nauseous."

And I thought I was cynical!

As for Livio's point, I think the main one is that the best job plan is to get elected at the end of a recession. Seems to work every time.


I think you are ignoring inter provincial migration? If Ontario were to generate those jobs, then people would move to Ontario (or wages would rise so much that employers wouldn't create those jobs).

Only if *all* Canadian regions generated that kind of jobs growth would we have a demographic problem.

If you consider programmes like that which provides Ontario farmers with fruit and vegetable pickers, ie importing seasonal contractual labour from the Carribean, then it's clear that the international labour market is also available to Ontario employers (although not always fully efficiently-- Federal immigration policy counts).

In addition another source of elastic labour supply is non working youth (in education/ higher education) and early retirees (plus extensions in healthy life that have taken place). If real wages start to rise, kids will cut short their education to join the workforce (given the cost of student tuition and the limited relevance of most undergraduate degrees to career, that's probably an increasing trend anyways). And those retired early (often not by choice) will come back to the workforce/ keep working.

The problem for Ontario will be getting that economic growth, not finding workers to do the work.

There, I am much more pessimistic. Ontario's economy seems to be, at least relatively, in long term decline-- somewhat disguised in recent years by the GTA property boom.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad