Why do people think that Canadian job creation rates require a policy response?

Pundits and politicians are calling on the government to "do something" about the state of the Canadian labour market; these calls are presumably based on a perception that job growth is still weak. It's not clear upon which data this assessment is based. The data that would be most informative are no longer being collected, and the best available proxy that I can find suggests that rates of job creation are back to pre-recession levels.

Here is a graph - taken from Calculated Risk - that I wish I were able to make for Canada:

 

This tells the story of how US employment fell and why it's struggling to recover. The drop in employment wasn't due to an increase in layoffs. Instead, the rate of hiring fell by some 20%, and it still hasn't recovered. The only reason why employment isn't falling further is that quit rates have fallen: more people are holding onto jobs that they might otherwise have left if there were better opportunities. So the problem in the US really is one of job creation.

When Statistics Canada releases its monthly Labour Force Survey, it publishes numbers for the stock of people who are employed and unemployed in a given month. This is of course a very important piece of information to have, but our understanding of the state of the labour market would be even better if we had data on the flows into and out of employment. There is no Canadian counterpart to the JOLTS - although it may be possible to tease some of this information from the micro data files.

A series that should be a good proxy for hires is the number of people who have been at their current job for a short period of time. The LFS provides numbers for those with various durations of job tenure, so I'm going to look at the shortest duration: those who have been at their current job for three months or less. If hiring rates move, this series should move as well (with a lag, of course).

The data are not seasonally adjusted. This matters, because much of the Canadian labour market is driven by seasonal factors. So the following bar charts group the data by month, so that deviations from the usual seasonal movements can be identified more easily.

 

The first thing to note is the scale of the vertical axis. In September 2011, 1.2m workers - 7% of the total - had been at their current jobs for less than three months. The scale of the gross job flows in and out of employment is one or two orders of magnitude greater than the net flows that make the headlines.

The monthly short-term tenure series started falling behind their usual levels in early 2009, and were significantly so from April 2009 through to January 2010. (Recall that this series reacts to changed in hiring with a lag of up to three months.) But since then, and certainly in 2011, the numbers have been roughly consistent with pre-recession levels.

The same narrative applies to the creation of full-time jobs:

 

I don't have much to say about part-time employment patterns, but for completeness, here is the graph:

 

My reading of the data of which I'm aware suggests that current rates of job creation are consistent with those observed during the last expansion, and have been so for a year or so. Calls for the government to "do something" are misplaced; the labour market has been functioning normally for quite some time now.

Comments

[Nick Rowe]

Neat! I always wondered why I never saw a JOLTS-equivalent graph for Canada.

I expect there are fixed costs of running a survey, if only in the sample size sense. The cost per head of population of getting data on a big enough sample to be useful would be lower in the US, just because it has a bigger population.

There's a Phillips curve, which shows inflationary pressures as a function of the *level* of unemployment. I think that there's an additional variable that belongs in that Phillips Curve: inflationary pressures depend on the rate at which unemployment is declining (employment is increasing), in addition to the level of unemployment. There's a "speed limit" not just in the long run rate of growth, but in how quickly we can get back to that long run growth path if we are below it. Hiring is costly and takes time.

[Determinant]

Well, a couple of things.

1) Quality matters. When one is a graduate, wants to start a family and become independent you want a full-time job in your sector. Otherwise you have to settle for something less. University graduates in Call centres for $12/hour comes to mind. This is underemployment. The other old dodge here is grad school which is also a form of underemployment.

2) Lack of hiring leads to long term unemployment. I can always tell the unemployed person at the party, they keep to themselves and don't make conversation because the REALLY, REALLY don't want to the conversation to turn to "So what you do for a living?" "I'm unemployed..." and the patronizing suggestions start from there. See 3 for what constitutes patronizing.

3) Duration of employment search from acknowledgement of submission to offer of employment. This often takes months and if one job doesn't happen then you have to start all over again. This really prolongs the pain of unemployment. It also makes EI's unemployment duration calculation a joke for many people.

It's a sad commentary on the private sector when the federal bureaucracy, the Public Service of Canada runs a better HR process (more e-mails saying where you are) and plain out moves through the stages faster than the private sector.

Also if you follow the "Call my cousin Vinnie at this place..." advice I described in (2) you'll wind up in this purgatory for months if you're lucky to even get an acknowledgement.

4) What's normal for the private sector? Here's a great article illustrating and explaining what I intuitively knew, that the private sector has a dearth of earth-shaking grab-you-by-the-collar I'll hire any warm body kind of investments that pay off. It explains why even in "good times" earlier in the decade jobs were hard to come by for many people.

http://innovationandgrowth.wordpress.com/2010/02/22/why-isnt-the-innovation-economy-creating-more-jobs-part-i/

It follows along from Nick's idea in that if you could have the entire private sector in a room and asked "OK, who's expanding and creating lots of job" everybody would look at the floor and shuffle their feet.

[Livio Di Matteo]

Hi Steve:
When I teach public finance, in the intro lecture, I mention that when faced with public clamor to solve a policy problem that may be real or perceived, there are five tools or approaches that government can use to address the request: tax, spend, regulate, borrow and ...At this point, I pause and ask the class what they think the fifth thing a government can do is. They generally provide endless variations on the previous four tools. Almost never do they state "Do nothing" though someone did once suggest study the issue further which comes close. Unfortunately, politics seems to require that governments always respond to an issue by trying to do doing something about it even if they do not fully understand the numbers behind the issue. Of course, the trick of good government is not doing nothing - it is to know when to do something and when to do nothing.

[Stephen Gordon]

That's brilliant!

[Mike Moffatt]

"It follows along from Nick's idea in that if you could have the entire private sector in a room and asked "OK, who's expanding and creating lots of job" everybody would look at the floor and shuffle their feet."

The funny thing is, *every* time I meet with fellow entrepreneurs, the first thing that comes up is how we're all trying to hire, but we can't find decent candidates.

How you see the problem is entirely on which side of the fence you're on. And our company, for one, is hiring.

[Determinant]

Just one thing, Livio. Remember than Keynesianism had the intellectual underpinning of "saving capitalism" at a time when the capitalist system was at best not fashionable, at worst it was under attack and being subverted.

The political strife of the Depression (e.g. the On to Ottawa Trek in Canada or the Bonus March in the US), the massive government-financed war economy and the subsequent post-war prosperity convinced the Western political and intellectual elite that in order to have a stable market economy in a Western electoral democracy, you needed to have full employment. Put another way business had to be seen as making every effort to attain full employment and people had to believe that business was distributing wealth widely and fairly. The corollary is that long-term unemployment had to banished by any means necessary, the formation of sizeable group of people who cannot access the market's benefits is poisonous to popular assent to a market economy.

No, I'm not a raving Marxist, I'm repeating the common historical narrative post WWII.

Business leaders, politicians and academics have not paid so much attention to the idea of selling people on capitalism, in particular since 1990 and the collapse of the Iron Curtain.

But the Tea Party and the Occupy Wall Street protests are the kind of direct, popular challenge to a market economy and the political order in which it operates that we haven't seen in a long time. If people follow these movements and stop believing in a market economy then we won't have a market economy.

Yes, it's politics. People's beliefs are politics. If business can't provide employment and government won't then people will believe in neither. That's the politics of the dispossessed. Do we as a society really want to go there?

[Determinant]

"It follows along from Nick's idea in that if you could have the entire private sector in a room and asked "OK, who's expanding and creating lots of job" everybody would look at the floor and shuffle their feet."

The funny thing is, *every* time I meet with fellow entrepreneurs, the first thing that comes up is how we're all trying to hire, but we can't find decent candidates.

How you see the problem is entirely on which side of the fence you're on. And our company, for one, is hiring.

OK, but I see your assertion and raise you a macroconomic graph detailing the woeful job creation record of private investment in the United States.

[Mike Moffatt]

Canada is not the United States.

[Determinant]

Canada is natural resources combined with the industrial base of Southern Ontario and Quebec. If you are a manufacturer in Southern Ontario you are tied in to the United States such that you may as well be another state. Been there, done that, know what it looks like. Besides the research sectors detailed in that article cross borders easily.

The oil sands are nice but given the high amount of energy needed to extract them and the consequent low productivity of energy production from them (too much energy input for too little extracted, an idea from a former geologist with the Geological Survey of Canada) it remains to be seen if they will ultimately be a good investment.

Second, I knew you'd rebut with "I can't get decent people", Mike. It's entirely compatible with an investment dearth. The actual returns from investment in the private sector are too low so the private sector is forced to cut costs. One way is to eliminate training. You do that by hiring experience and not training yourself. It can work for a firm but it's a zero-sum game for the country.

Flip it around. If we were experiencing an investment boom with high returns, you and most firms would hire any warm body they could get their hands on. That is exactly what happened in Canada in the 1940's. It's not happening today.

[Mike Moffatt]

We are hiring, though. We've hired people and have continued to hire people. You can believe my anecdotes or you can believe the Canadian data Stephen posted. Both show the same thing.

When we say 'good people', it's all personal habits, not experience. We hire people right out of school. No experience necessary or even desired.

Again, it's a simple 'grass is greener' situation. You see the problems being on your side of the table. I see the problems being on my side of the table.

That's why anecdotes aren't particularly illustrative - you need to look at the data. The *Canadian* data.

[Mike Moffatt]

"Canada is natural resources combined with the industrial base of Southern Ontario and Quebec."

These two industries make up less than 15% of Canadian employment. See: Employment by industry and sex.

[Jim Rootham]

@Mike

How do you treat the people you do not hire?

[hosertohoosier]

I think you have a few things at play.

1. Canadians watch a lot of American television, and implicitly adopt American memes. Even though Canada's economy has recovered, Canadians may not have entirely gotten that picture yet (although we sure are re-electing incumbents).

2. People don't always get their picture of how the economy is doing from job reports - in fact very few do. Other than scuttlebutt, the stock market is probably the more visible indicator of how the economy is doing, particularly for people that invest. Not only is the TSX well below its pre-crisis peak, it has fallen considerably in recent months.

3. Aggregate statistics may mask stark sectoral differences. If you look at employment by industry, many have still not recovered (as of 2010: http://www40.statcan.gc.ca/l01/cst01/econ40-eng.htm).

Change in employment in selected industries, 2008-2010
Manufacturing -11%
Farming: -7%
Transportation: -5%
Education: +2.7%
Healthcare: +7.3%

And I suspect economists understate the costs of this kind of transitional employment. Even if workers in the manufacturing sector ultimately find jobs elsewhere, it often means moving (breaking up social ties), and giving up on human capital they developed over the years. Nor should economists assume they are immune to these kinds of effects. In my discipline, political scientist, a whole crop of scholars (Sovietologists) found themselves obsolete when the Berlin wall fell. How soon is it till Watson can do your job?

[Determinant]

@Mike

The Canadian data ARE in doubt. Big doubt. Let me introduce you to the Peterborough Problem. I know Peterborough, ON well. It has an 11% unemployment rate.

http://www.thepeterboroughexaminer.com/ArticleDisplay.aspx?e=3314393

Statistics Canada reported that 3000 manufacturing jobs had been created in Peterborough from July 2010 to July 2011. This mystifies everyone. They would like to know where they are. The newspaper and the rest of the media, the Mayor and Council and the Economic Development Corporation would all like to know where they are so that they can throw a party for the employers responsible. The local newspaper, the Examiner, called around asking HR departments about their hiring. Nobody's talking, nobody's bragging and this they brag about. You can't hide 3000 cars going to jobs every day.

So instead of a party we are left with the fact that Statistics Canada has a rogue survey in Peterborough. Nobody is shooting the lights out.

Secondly Mike, does your HR have a filing cabinet full of resumes? I bet they do. If this were an investment boom and we had full employment, you would not have that cabinet. You would be begging for people, you couldn't afford to be picky and you would be complaining that wages were going through the roof.

Instead you described being picky, waiting for the "right person" and you are not facing wage pressure. This is not a full employment picture not even a strong employment picture.

[Determinant]

Thanks for that entertaining fact that Statscan can't spell labour (not labor).

That survey also lists the following:

Trade,
Transportation and warehousing,
Professional, scientific and technical services,
Business, building and other support services

That falls under "manufacturing" in my book. The trucking firm that hauls goods and the support firm that calibrates meters are all dependent on end manufacturing firms for their survival. This is why their offices are located right next door.

[Jacques René Giguère]

Mike: Full employment looks like the Québec North Shore. In Baie-Comeau, one of the largest and busuest restaurant closed two weeks ago. No cooking staff. All went to the mines and hydro camps. In Sept-îles, McDonald's close at 2100 ( 9 anglo time). No staff. I recall, in the late 70's the police closing the bars at 3 am in this very rowdy town ( we still have the highest ratio of cops per capita in the province) but closing on a full crowd at 9 in the evening because there is no staff? That's weird.
Boys are going back to school because mines have promised jobs to anyone graduating in Elecric or Pneumatic control and Mineral Technology ( whether the jobs will be there in 3 years is better left for another day).
My college train people for mines in the South Pacific.The company there wanted some of our staff to accompany the students and help them fit in their new jobs. Dean and bord were delighted. Dean of study office refused at first, collective agreement , dean and board notwhithstanding, because they knew it would be almost impossible to find replacement.
Basement rooms are advertised as having " almost no molds".
That's what full employment is like.

The lack of trained staff is such,that if your business was located here, you would hire me as econometrician...

[Paul Friesen]

I don't think we have a shortage of jobs in Canada now. Certainly there is no evidence for it in the number of "help wanted" signs I see around Guelph (although I understand that Guelph is doing better than most of the country).

However, I would argue that the Canadian situation could best be described as a full employment liquidity trap. It's still a liquidity trap because interest rates are still essentially at the zero lower bound. This matters, because it means that normal monetary policy won't work if we get a negative demand shock like a drop in exports. This seems all too likely given the political and economic conditions in both the U.S. and Europe. I hope that the government is prepared to step in with more fiscal stimulus if that happens.

[Nick Rowe]

Jacques: that sounds more like "over-full employment" to me. Excess demand for labour. Wow!

Think I will postpone my tour of the North Shore. Wait till they need the extra demand.

[Nick Rowe]

Paul: I think I disagree on the "liquidity trap" bit of what you said. And it's not just that the Bank of Canada can cut at least another 75bp of the current rate. It can repeat what it did a couple of years back, and provide guidance about future interest rates, if needed. It can do asset purchases if needed. Compared to the Fed, the BoC has shown a greater willingness to do whatever it takes to stick to a *symmetric* inflation target (i.e. 2% doesn't mean "less than or equal to 2%"). Plus, it has built up some more credibility over the last couple of years, I think, by being seen ready to do whatever it takes.

Hope I'm right, anyway. But sure, I would still like to keep fiscal policy in my back pocket as an insurance policy, in case I might be wrong.

[Hans Blix]

Shouldn't this account somehow for a (trend) growing labor force/population?

Looking at the "less than three months" graph, if the dark brown spike were as tall as the light blue spike, that would just signify that the same number were employed as four years ago. But presumably the labor force (strictly or loosely defined) is larger now? Put differently, what's the counterfactual for a healthy job market?

[Stephen Gordon]

Surprisingly, no. Since the change in the number of those employed is the difference between hires and separations, you get employment growth so long as one is bigger than the other. You don't need a trend in either.

Same thing with housing starts. You don't need a rising number of housing starts with a growing population; you just need it to continue at a level that's greater than the number of houses lost.

[vimothy]

[Hans Brix? Oh no!]

Oh, herro. Great to see you again, Hans!

[jesse]

I would take a step back and ask where previous employment gains came from. From what I see it's based on government and household debt levels increasing. The latter is now to a point where adding debt is net destructive so it's worthy asking where the next round of job gains are going to come from. At current growth rates we are still years away from full employment.

Another interesting stat to track for Canada is the % of long-term unemployed. In the US it looks terrible, a long-term liability for the government the longer people stay out of work.

I also wouldn't use the previous recession's recovery as indication of a "normal" recovery. Look at previous recessions from the past 50 to 100 years to gain some perspective; 2001-2006 was far from normal.

[Hans Blix]

"you get employment growth so long as one is bigger than the other."

No, the question is if this growth rate is large enough to keep up with population growth. If job growth is (proportionally) less than population growth over the entire period, isn't this a problem?

Thanks vimothy, great to be here.

[Hans Blix]

Ooops. Sorry, I recognize my error. So the issue is that job creation is back to it's level trend. I have a quibble about whether we should see job growth bounce out of a recession rather than attain prior rates but that is another issue. Sorry, my misinterpretation of the data.

So the appropriate statement given the data is "the number of people who are new (tenure<3 months) at their job now is similar to the number of people who were new at their job in 2006."

hmmm, but this interpretation really hinges on this being a good proxy for hires and not something that trends in any way over time. It would be interesting to see how this behaves in normal times (say since 1990) to rule that out.

Thanks Stephen.

[Bob Smith]

"Aggregate statistics may mask stark sectoral differences."

Fair point, but itreally only modifies Stephen's point. If the issue is one sectoral shift, than the appropriate policy response will be different than what it would be if the issue was an absence of job creation. So the calls for a job creation strategy are still misguided.

Hoosier,

If the problem is a job creation problem (understood as being a situation where employers are hiring fewer people than they would if they perceived the economy to be more robust than it is), than an employment strategy (say, subsidizing the hiring of new employees in the manufacturing sector) might make sense. Subsidize employment for two years while the economy recovers (hopefully?) and, with luck, the employees stay on as permanent employess.

But, if the problem is one of a sectoral shift, than that isn't a particularly good strategy. Sure, the government can subsidize employment in the manufacturing sector for two years, but if employers are shifting to China anyhow, that just defers the problem for two years, once the subsidy ends, the jobs will too. In that case, the better use of public funds would be to help those employees transition into new jobs in different sectors (for example, training, relocation allowances, bridging payments for retirement, etc.), rather than stringing them along for a couple of years.

While either scenario might merit a policy response, I think Stephen's overarching point is that we should make sure that the response reflects the real underlying problem.