Three Changes Making It Difficult To Find a Low Skill Job

In the comments of an earlier post we discussed the difficulty entry-level applicants are having in obtaining jobs.  In this post I discuss 3 changes from the last 12 months making it difficult for workers with low- to no-skills find employment.

When an employer has a low-skill task that needs to be done - whether it be low level data entry or manual labour, she has a number of alternatives (or substitutes) to hiring a low-skilled worker.  She could:

1. Find a technological solution. Rather than hiring another worker, is there a piece of equipment or machinery that will do the job, or make existing workers more productive so another person is not needed?  The automated checkouts in grocery stores might be a good example, though I've often wondered how much labour they truly save.
2. Outsource/offshore/subcontract. Instead of doing the work in-house, can we pay someone else to do it?  Off-shoring has become easier than ever for even the smallest of businesses thanks to services such as eLance and Mechanical Turk.
3. Do without. Instead of having 8 cash-registers open, can we get away with having only 7?

Naturally not each substitute is applicable to each situation, but in many cases 1 or 2 would apply.

So what are three changes?  I'll go through them from (arguably) least to most important:

1. Employer Side EI Premium Hike

Overall a small-hike for employers: 7 cents per $100 starting in January 1, 2011.  Over time this usually ends up being paid by workers (through lower wages), though due to the presence of a price-floor (minimum wage) this more likely manifests itself in a reduction in jobs.  For a low-income earner making $10,000 in a year an employer will pay $249.20 in additional EI premiums, up $7 from 2010's $242.20 amount.  The employer will also pay $321.75 in CPP contributions, which is unchanged from 2010.  The small increase in EI premiums should have a very slight, but negative, effect on jobs. 

2. Rising Canadian Dollar

Consider the case of outsourcing.  When a Canadian company outsources to India or Bulgaria or Indonesia, typically the contract is in U.S. dollars.  A company has a data entry project, with the alternatives being hire someone or outsource.  In early 2009 when the Canadian dollar was at 80 cents, a $2000 USD price quote for outsourcing equaled $2500 Canadian.  Today at parity, that same quote now is $2000 (20% less), making the outsourcing option far more attractive.

3. Increases in the Minimum Wage

In Ontario, the (nominal) minimum wage has risen 50% since 2004, from $6.85/hr to $10.25/hr (source), including a 75 cent hike in 2010.  Since it is possible on eLance and Mechanical Turk to hire workers in say, the Philippines or Bulgaria for $3-4/hr, domestic workers face a tough time competing for jobs which can be offshored. That's not to suggest that the minimum wage is necessarily bad for the low-skilled.  I suspect most low-skilled workers would rather live in a world with a $10.25/hr minimum wage where it's harder to find worker than a $6.85/hr one where it is easier.  But given the possible alternatives (technology, outsource, do without) a higher minimum wage does reduce low-skill employment.

Outlook for the Future

A strengthened domestic economy will certainly assist low-skilled labour.  But they won't be helped by EI premiums that are set to rise again in 2012 and a Canadian dollar that continues to strengthen.

Comments

[Robert McClelland]

a higher minimum wage does reduce low-skill employment.

Only in your textbooks. In the real world though, not at all.

[Mike Moffatt]

Yeah, I must have misread all those studies. And all those payments I've been sending to India? All in my head.

[Robert McClelland]

Is nuh-uh really your response? Show me the jobs that have been lost due to any of the hundreds of increases in minimum wage that have occurred just in Canada. Hell, at this point I'll accept even one lost job attributable to a rise in minimum wage as evidence.

[Mike Moffatt]

How about the guy I fired?

[Mike Moffatt]

Specifically - was paying a guy to do web work for us. Minimum wage kept going up. Sent the work to India instead. There. One job. What do I win?

[Stephen Gordon]

And if anecdotal evidence isn't enough, you can read - or have translated - Pierre Fortin's survey of the empirical literature.

[Frances Woolley]

Mike - the title of the post is "low skill job" - and the example of jobs being killed off by the rising minimum wage is web work?

Scary.

[Lord]

Forgot higher real energy costs make more uneconomic. They aren't back to 147 but still far higher than the 30-50 of old.

[Mike Moffatt]

"Mike - the title of the post is "low skill job" - and the example of jobs being killed off by the rising minimum wage is web work?"

Low level, data entry stuff. But I'm sure if I posted a job @ minimum wage for very basic HTML coding I could have 100 applicants within 3 weeks.

[Robert McClelland]

What do I win?

The Cheap Bastard of the Year Award.

Pierre Fortin's survey of the empirical literature.

Nice wild goose chase, Stephen.

[Laura S.]

The automated checkouts in grocery stores might be a good example, though I've often wondered how much labour they truly save.

Here in California, the situation might surprise you. A typical large grocery store will have a few lanes open with cashiers, but there are then also a half-dozen guys or more in the back, "waiting". The union work rules at the two chains I'm familiar with say that the lines have to be X long for Y minutes, and then one more cashier can be called up. Then, the process repeats. As soon as the lines clear, all the extra guys rotate back to waiting in the back.

So this is a very slow ratchet. Sure you hire a lot of guys, but very few actually do any work for the hours they are clocked in--they goof off in the back. Hiring one more guy is a very ineffective solution because of the work rules--but its impossible to change the work-rules. We went through a two month strike. The managers ran the checkout almost as well--no breaks in back, but the union won the public relationships and customers started boycotting the stores. So the work-rules stay.

Enter automated checkouts. The customers like them because they clear through faster--lots of machines ready to go counterbalances the slightly slower process and the hassle of self-checkout. Best of all, one cashier supervises all the machines and is always working--not in the back goofing around. They can be paid a better wage without strangling the business.

FYI, other chains without these work-rules are insurgent and are grabbing market share gradually. I'm thinking of the German run Trader Joes and the wholesale chain Costco, but also a lot more small operators with only a dozen stores or less. These places staff all cash registers, all the time, and the cashiers move fast. The prices are lower and the customer service experience is higher.

The labor costs have gotten so bad that the pricing advantage of the large chains on the products isn't enough to compensate and exclude the upstarts.

[Mike Moffatt]

Thanks for posting that link, Stephen. One thing I love about Google Chrome is how it automatically translates webpages.

RE: Minimum wage costing jobs. Think of call-centers, data entry, basic HTML coding.

If the minimum wage were still $6.85, the hourly cost of having an employee would be $7.36/hr at 2011's CPP + EI rates. But now it's $11.01 ($10.25 + CPP = EI). (And that's not including any other expense such as office space, etc.)

I can get someone in India to do the same job for $3-4/hr. All else equal I'd prefer to have the Canadian, as there is a productivity loss (language and cultural barriers, different skills), monitoring costs. Even the time-zone is a bit of a pain.

There's some point at which I'm indifferent between hiring the person in India or hiring a Canadian. It's certainly more than $4. But to assume that point of indifference cannot possibly be between $7.36 and $11.01 is just weird.

That's why so many businesses outsource these to India - because for many tasks it just doesn't make economic sense to pay someone $11/hr here compared to $3-4/hr there.

[Mike Moffatt]

"The Cheap Bastard of the Year Award."

Flattery will get you nowhere with this Scotsman.

[Robert McClelland]

If the minimum wage were still $6.85...

...you could still hire someone in India to do the job for $3-$4 less. This is not an effect of increases to the minimum wage but an effect of cheaper labour in another region. And that labour will always be cheaper somewhere in the developing and undeveloped parts of the world regardless of whether or not the minimum wage rises here.

[Mike Moffatt]

"...you could still hire someone in India to do the job for $3-$4 less. This is not an effect of increases to the minimum wage but an effect of cheaper labour in another region. And that labour will always be cheaper somewhere in the developing and undeveloped parts of the world regardless of whether or not the minimum wage rises here."

Right, but there are additional costs, as I pointed out: language and cultural barriers, different skills), monitoring costs. Even the time-zone is a bit of a pain.

Now, it might be the point of indifference is below $7.36, in which case, you're right the rise in minimum wage did nothing. But to assume that point of indifference cannot possibly be between $7.36 and $11.01 is just weird. In our case, it was the point of indifference. If we could hire someone locally for $7.36, we would, because it'd get rid of a bunch of headaches. But for $11.01/hr? We'll take some aspirin for our headaches instead.

[Stephen Gordon]

Nice wild goose chase

My mistake. I had mistaken your request for evidence as a request for evidence.

[Adam]

"I suspect most low-skilled workers would rather live in a world with a $10.25/hr minimum wage where it's harder to find worker than a $6.85/hr one where it is easier."

Assuming that this is true, I doubt that it's of much comfort to those who can't find a low-skilled job to know that the majority is OK with their unemployment.

[Adam]

"The Cheap Bastard of the Year Award."

You're free to hire the fellow to do your blog. I'm sure that Mr. Moffat would be happy to give you his contact info.

[Mike Moffatt]

"Assuming that this is true, I doubt that it's of much comfort to those who can't find a low-skilled job to know that the majority is OK with their unemployment."

It really isn't. And it's just a guess on my part. But I can't picture a lot of low-income/low-skill workers screaming 'OH NO!' whenever the minimum wage is raised.

[Jon]

Its definitely been unfortunate that in the middle of a major recession and low inflation, the minimum wage has risen not once but three times in the US--its up forty percent since 2007.

[Patrick]

Outsourcing isn't free. My observation is that it's rather expensive and time consuming managing outsourced/off-shored projects, and the productivity and quality almost never stack-up to what you'd get in Canada. Admittedly, my experience is with software and electronics, so it's not exactly low skill.

I think the firms management got sucked into the promise of lower wages but didn't figure in the additional costs of managing a project from halfway around the world in a place with a different work ethic etc. They discovered through bitter experience that it didn't really cost less overall, and in one case repatriated the work and in another had to get the Canadian engineers to handhold the offshore contractor through the project. My guess is that they couldn't pull the plug on the project because too many big egos where on the line and admitting it was a mistake would have implied heads rolling, so the CYA maneuvering started (blaming the contractor, blaming their underlings, etc).

[Robert McClelland]

I had mistaken your request for evidence as a request for evidence.

Your link didn't lead to evidence. It led to someone referring to evidence. For example;

Canadian work that marked the last decade are mainly those of Baker, Gunderson and colleagues (Baker, Benjamin and Stanger, 1999, Baker 2005; Campolieti, Gunderson and Riddell, 2006). According to the review by Gunderson (2005, 2007)...

[Robert McClelland]

In our case, it was the point of indifference.

But that point of indifference also existed before the rise in minimum wage and its doubtful that it changed much since wages even in areas of cheap labour are not static. So labour that costs $7 here and $4 in India is no different than labour that costs $11 here and $8 in India.

[Stephen Gordon]

There's a bibliography that you can use to look up the original papers.

[Robert McClelland]

Yes there is, Stephen. And if I were a rube I'd surely make use it to prove your point.

[Stephen Gordon]

Okay, seriously: you have officially removed yourself from the Evidence-Based Community. You aren't interested in statistical analysis, data analysis, or even taking the trouble to read the work of those who are. Next time you feel like ranting about ignorant, know-nothing right-wingers, remind yourself of this thread.

[Mike Moffatt]

"Outsourcing isn't free. My observation is that it's rather expensive and time consuming managing outsourced/off-shored projects, and the productivity and quality almost never stack-up to what you'd get in Canada."

We've found the same thing. My rule of thumb is that if you don't get at least a 50% cost savings, it's not worth it due to the added administrative headaches. But that % would be different for different situations.

"But that point of indifference also existed before the rise in minimum wage and its doubtful that it changed much since wages even in areas of cheap labour are not static. So labour that costs $7 here and $4 in India is no different than labour that costs $11 here and $8 in India."

Agreed. But our outsourcing costs haven't doubled in the last 5 years. Particularly when you consider that the rising Canadian dollar has given us a 25-30% cost advantage from forex alone (see point 2). If anything, outsourcing when priced in CAD is cheaper than it was 5 years ago, as the rise in wages in India has been less (in percentage terms) than the rise in the CAD vis-a-vis the USD.

[Robert McClelland]

You can't handle the evidence.

In 2004 there were 621,000 minimum wage jobs in Canada. In 2008 there were 751,400 minimum wage jobs in Canada.

Did any of those studies explain why, despite numerous minimum wage increases over those 4 years, the number of minimum wage jobs increased by 21%?

[Mike Moffatt]

"Did any of those studies explain why, despite numerous minimum wage increases over those 4 years, the number of minimum wage jobs increased by 21%?"

Isn't that what you would expect with a minimum wage rise, as a whole host of $8 and $9 jobs that weren't minimum wage before are now minimum wage?

That's why you look at studies instead of just grab single pieces of data - to help you control for such factors.

[Stephen Gordon]

You're making the post hoc ergo propter hoc mistake. Careful statistical studies - the sort you apparently refuse to look at - do not.

[Mike Moffatt]

I don't want to alarm anyone, but it was 25 degrees warmer 6 months ago than it is now. If this holds up, we'll all be dead by June.

[Patrick]

"Did any of those studies explain why, despite numerous minimum wage increases over those 4 years, the number of minimum wage jobs increased by 21%?"

From Google:

GDP 2004: 992 billion
GDP 2008: 1.499 trillion
% change: 51%

So you're wondering why there are so many fewer min wage jobs than one might expect at first glance?

[Robert McClelland]

Isn't that what you would expect with a minimum wage rise, as a whole host of $8 and $9 jobs that weren't minimum wage before are now minimum wage?

If that were the case then why doesn't it hold up over time (see figure 3) despite a steady rise in minimum wage (see figure 2) over the same time period.

[Mike Moffatt]

Then why doesn't what hold up?

Back in the days when I was very active with the Green Party, I spent a lot of time debating far right climate change deniers, who would point to a snowstorm and say 'where's your global warming now?' Single data points are no match for controlledstudies. This whole conversation reminds me of those days.

[Robert McClelland]

So you're wondering why there are so many fewer min wage jobs than one might expect at first glance?

Not at all. We simply reached the saturation point for employment (ie. there were no more people left to work for minimum wage).

[Robert McClelland]

I've just gone from 4 data points (2004-2008) to 11 (1997-2008). How many do you require?

[Robert McClelland]

I don't refuse to look at statistical studies, Stephen. I refuse to spend my time googling in order to find the statistical studies that provide the evidence to support your point.

[Stephen Gordon]

At some point, you have to do your own homework. That's part of the burden of being a member of the Evidence-Based Community.

[Robert McClelland]

At some point, you have to do your own homework.

Agreed. That's why my links go exactly to the evidence that back up my point instead of leaving you guessing about what you're supposed to look at.

[Patrick]

"Not at all. We simply reached the saturation point for employment (ie. there were no more people left to work for minimum wage)."

You lost me on with that one.

[Nick Rowe]

Suppose workers are different. They vary by some attribute, call it "skill" (or "employability"), that employers want. There is a distribution of skill, with a single peak mode, with the frequency falling each side of that mode, so only a few workers have very low or very high skill. With no minimum wage laws, we would see a distribution of wages similar to the distribution of skill.

Assume the demand curve for labour at each level of skill slopes down.

Starting with a minimum wage of $0, what happens as we slowly raise the minimum wage, holding everything else constant? In particular, what happens to the number of workers employed at the minimum wage?

There are two effects:

1. The downward slope of the labour demand curve reduces the employment of workers at each level of skill for which the minimum wage is binding.

2. The upward slope of the skill distribution increases the number of workers for whom the minimum wage is binding.

So those two effects work in opposite directions.

If we start from a very low minimum wage, which is binding for almost no workers, the second effect must initially be bigger than the first effect.

Therefore Robert's observation, that increasing the minimum wage results in more workers employed at the minimum wage, is consistent with a model in which increases in the minimum wage reduce employment. And that is even if we hold everything else constant.

[Robert McClelland]

You lost me on with that one.

No doubt. One day I should brush up on my econ lingo so I'm talking your language. Anyway, if you have a pool of 10 workers and all 10 are working you've reached the saturation point. Regardless of the demand for more workers, no new jobs will be created from that point on simply because there are no more workers to fill them.

By 2008 we were at or very near the saturation point. The employment rate was a record high of almost 64% and during the few years prior to that bringing in temporary workers had become an issue. Therefore no more additional jobs (those in excess of the turnover rate) could be created simply because there were no more workers to fill them.

[Robert McClelland]

Therefore Robert's observation, that increasing the minimum wage results in more workers employed at the minimum wage, is consistent with a model in which increases in the minimum wage reduce employment.

Except Robert's observation was a trap designed to elicit this very response. When you look at the longer time frame (figure 3) the number of minimum wage jobs is fairly static over that 10 year time period which is not consistent with the model in which increases in the minimum wage reduce employment.

And this is the core of my argument; that the theories and models, while they may work exceptionally well in theory, fall apart in reality. There simply is no evidence to support the claim that increases in minimum wage reduce employment.

[Mike Moffatt]

"I've just gone from 4 data points (2004-2008) to 11 (1997-2008). How many do you require?"

You've missed the point - the raw data doesn't tell you much of anything. You need to control for all the confounding factors. It's no different than someone arguing against anthropogenic climate change by posting temperatures from 11 different dates. Even if they did provide links from the Weather Network.

[Adam]

Will someone please just ask RM why, if minimum wage hikes have no effect on unemployment, we don't just jack it up to $50/hr and make everyone rich?

Never mind, I think I just did.

[Nick Rowe]

Robert: You cited some data on the number of workers employed at minimum wage, and said it was evidence against the theory that an increase in minimum wages causes a reduction in employment.

I showed that *even if you ignore Mike and Stephen's point about other things changing over time* your data does not support your conclusion. It is possible for an increase in the minimum wage, other things constant, to decrease total employment, but either increase or decrease the number of people employed at the minimum wage. Because there are two effects going in opposite directions.

Now you say the data is the opposite of what you originally said it was. So what? My point was that the theory that an increase in minimum wages causes a reduction in employment is consistent with that data moving in either direction.

I tried to engage you in a genuine discussion of the evidence. Now you say you were playing games with "traps". The fact that your "trap" failed is less relevant to me than the fact you seem uninterested in a genuine discussion of the evidence.

[Patrick]

"No doubt. One day I should brush up on my econ lingo so I'm talking your language"

Robert, I'm not an economist. I'm a software developer by training and trade. All you need to do to communicate effectively with me is make sense.

[Robert McClelland]

Now you say you were playing games with "traps".

Yes, because there's one thing I've learned you have to do when having a discussion with an economist; drag out every one of their canards and beat them silly. Until that is done the discussion cannot progress. That was one canard and I still have to drag the phantom job losses canard (ie. without the minimum wage increase we're making a random guess that there would have been more jobs created) out of the closet.

It is possible for an increase in the minimum wage, other things constant, to decrease total employment, but either increase or decrease the number of people employed at the minimum wage. Because there are two effects going in opposite directions.

Am I reading this right. If the number of minimum wage jobs increase that is consistent with your theory and if the number of minimum jobs decrease that is also consistent with your theory? No wonder economists believe this nonsense. They've convinced themselves that all the evidence supports their theory.

[Robert McClelland]

Will someone please just ask RM why, if minimum wage hikes have no effect on unemployment, we don't just jack it up to $50/hr and make everyone rich?

Because we don't live in Sillyland.

[Robert McClelland]

the raw data doesn't tell you much of anything. You need to control for all the confounding factors.

Only as it pertains to a theoretical discussion. When you're looking at reality however, all factors add up to paint an accurate portrait of what is really happening. And the accurate portrait of what is happening is no job losses. In reality job losses only occur whenever we enter a period of recession. At all other times the number of jobs increase regardless of what is occurring with those confounding factors.

[D. I. Harris]

Robert, you constantly state that raising the minimum wage will not engender job losses. Thus, the $50 minimum wage is a real question: if job losses only happen in recessions, a $50 minimum wage will not cause any job losses.

http://en.wikipedia.org/wiki/Reductio_ad_absurdum

[Robert McClelland]

Thus, the $50 minimum wage is a real question

Raising the minimum wage to $50 dollars would cause other real problems; such as inflation of unimaginable proportions.

[Nick Rowe]

Robert:

If the minimum wage were low enough, there would be no minimum wage jobs. Because all jobs would pay more than the minimum wage anyway.

And if the minimum wage were high enough, all jobs (if any remained) would be minimum wage jobs.

So, even holding everything else constant, data showing that the number of people employed at the minimum wage either increased or decreased would neither confirm nor disconfirm the theory that increases in the minimum wage reduced employment. (Nor would they either confirm or disconfirm the theory that increasing the minimum wage caused employment to increase.

Yep. Some facts are relevant for testing theories, and other facts aren't.

[Stephen Gordon]

Robert, I do hope you're actually paying attention here. Playing games with rhetorical tricks and traps really is no substitute for understanding.

[Robert McClelland]

Playing games with rhetorical tricks and traps really is no substitute for understanding.

I agree. Which is why you guys should stop doing it instead of bringing up absurd questions about $50 minimum wage levels or alternate realities in which the minimum wage is so low that nobody works for it. How about we agree to stick to what is actually happening in the real world for this discussion?

[Robert McClelland]

So, even holding everything else constant...

We do not live in a textbook filled with theories where everything else remains constant. We live in a real world where for example, the cost of reasonable minimum wage increases are passed on to the consumer who doesn't even notice thereby resulting in no lost jobs.

[Stephen Gordon]

You really should take the time to look up and read those econometric studies. They are careful to control for those other factors.

[Robert McClelland]

You should stop trying to send me on wild goose chases and just post a link to one that doesn't require me to be a member of something or other in order to read it.

[Larry Motuz]

In the real world, people have to eat. There is no minimum wage where that changes nor any upper bound to that need. One of the problems with economy theory is the idea that needs may be expressed as preferences with prices making one's final choices a matter of indifference choices relative to other goods.

With respect to the minimum wage, so many factors are at work that it is difficult to impossible to say, one way or another, that and increase/decrease in employment results. Higher wages, for example, reduce turnover and associated turnover costs. This lowers vacancy rates. Higher wages can lead to reduced hours. It can also lead to increased labour market participation, i.e., more people entering the work force. Other incentives, through tax policy, may allow employers to increase workers despite a hike in MWs.

One thing I've noticed about the comments is that no one has compared employment 'growth' per se with the employment 'growth' of minimum wage earners. Is it faster or lower? A simple shift and share analysis could help here.

There are many contradictory studies 'out there'. Seldom do they control for the same variables, and, even more seldom are they comparable.

[Stephen Gordon]

Sorry, Robert. You've long since burned away the stock of goodwill that would have justified doing your research for you.

[Robert McClelland]

Okay economists, answer this. Are there any comparable studies that show wage increases for everyone else not making minimum wage leads to job losses? I've never heard of one--not that this means there aren't any--but it's a little strange that economists only obsess over the employment effects from wage increases to the smallest and lowest paid percentage of the workforce. I mean, with the exorbitant wage increases CEOs receive for instance, theory would dictate that their jobs should be disappearing at an alarming rate.

[Andrew F]

"I mean, with the exorbitant wage increases CEOs receive for instance, theory would dictate that their jobs should be disappearing at an alarming rate."

Those pay rates are at least theoretically market rates, and not artificial price floors. Non sequitur.

[Frances Woolley]

I find this entire minimum wage discussion a bit frustrating.

Governments aren't stupid.

A minimum wage is a widespread social institution.

If there was an alternative to minimum wages that achieved the same or better outcomes, wouldn't we expect to see the minimum wage being replaced by this other institution?

I don't see that happening - every province has a minimum wage. No province is proposing abandoning its minimum wage.

Yes, I see federal and provincial governments introducing cash benefits that supplement minimum wage employment, particularly ones that take account of the special needs of parents, etc. But those programs create all sorts of inefficiencies too.

Once again, one of Nick's all-time greatest posts is worth re-reading: http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/02/of-horses-and-men.html

The irony of all of this discussion, of course, is that many of these so-called low-skill jobs are being done by university graduates, while high skill work, such as blogging or writing in the Globe and Mail's economy lab, isn't paid at all.

[Robert McClelland]

If there was an alternative to minimum wages that achieved the same or better outcomes, wouldn't we expect to see the minimum wage being replaced by this other institution?

The people who rail against minimum wage don't want a replacement. According to them, life was just grand for everyone when no minimum wage existed so we should just go back to having none at all and let the free market dictate how many cents per day the working scum should receive.

[Mike Moffatt]

There's a enormous difference between recognizing that increasing the minimum wage has a cost in jobs and calling for its abolishment. You'll note that nowhere in this thread do I (or I believe anyone else) advocate the latter.

If we're going to abolish EVERYTHING that has an opportunity cost, we're going to be very, very busy.

[Robert McClelland]

Sorry, that last comment was directed at clownservatives. They rail against minimum wage. Economists just obsess over it.

By the way, I plugged these search terms into google: effects wage increases employment. Every single one of the first 5000 entries were in reference to minimum wage increases. This leads me to tentatively conclude that only wage increases to the lowest paid workers in our world possess job destroying--and if you listen to clownservatives, economy destroying--powers.

[Stephen Gordon]

Try variations on 'estimates/estimated' 'labor/labour' 'demand' 'elasticity/elasticities'. There's lots of work there; I wrote some of it myself. Labour demand curves slope down; it's a general result. The only reason you see it more in the minimum wage file is that it's one of the few wages that is set by policy.

[Melissa]

I'm usually a lurker here, but I just have to offer up a scholarly book citation here. The research is done in the US, not Canada, but I would expect the same basic economic principles apply:

http://press.princeton.edu/titles/5632.html
Card & Krueger, "Myth and Measurement", Princeton Univ Press, 1997. This book pretty solidly debunks the claim that increasing minimum wage increases unemployment (at least in the kind of increments seen in the US, not the $50 jump mentioned in this thread).

It is cited in a nicely written op-ed by a member of a serious economic think-tank in the US, which gives a readable layperson's summary of the anti- side of the debate:
http://www.cepr.net/index.php/op-eds-&-columns/op-eds-&-columns/minimum-wage-myths

I would be very interested, and I mean that most sincerely, in any analysis of why the US might be so different from Canada that the data and arguments in those sources wouldn't apply here.

[Nick Rowe]

Melissa: "I would be very interested, and I mean that most sincerely, in any analysis of why the US might be so different from Canada that the data and arguments in those sources wouldn't apply here."

This is not my area. I am not familiar with the literature. But here is my *guess*:

There is one economic model that predicts that a *small* increase in minimum wages can increase employment. That's the monopsony model. Essentially, if employers have enough bargaining power they can push wages below the competitive equilibrium, so that there's an excess demand for labour. One of the predictions of that model is that employers will always want to hire more workers at the existing wage, but can't find any (there are "help wanted" signs everywhere, and any worker who wanted a job would have several job offers).

If you put a minimum wage on a monopsonistic labour market, *as long as you don't raise the wage above the competitive equilibrium*, the model predicts that employment will increase. Because higher wages increase the supply of labour.

So when I read the Card & Krueger paper, my first thought was "Hmmm. Maybe the US labour market for some low-skilled jobs was a monopsonistic market". And, IIRC, the US labour market around that time did seem to roughly fit the picture of a labour shortage that you would expect with the monopsony model. And certainly more so than Canada, which generally doesn't seem to fit that model. Maybe because Canada's minimum wages are already high enough? Maybe other differences?

Or, it may just be that Card and Krueger were wrong, or missed some other change that happened at the same time. Again, I'm not up on the recent research, but I think it all points in the other direction.

[Stephen Gordon]

Again, the Fortin survey I linked to above covers this.

The Card-Krueger results picked up only a short-run effect from a small increase in a minimum wage that was very low to begin with. If you look at the long run, the effect is stronger. If you look at increases in minimum wages that are relatively high (say, 45% of the average), then the effect is stronger. Other studies with Card-Krueger features - that is, measuring the short-run effects of an increase in a low minimum wage - get similar results, so their results needn't be discarded entirely. But they do need to be put into perspective. The Card-Krueger results shouldn't be generalised beyond the context that generated them.

[Patrick]

As Frances points out, it may be that the alternatives to a minimum wage are worse. Though, I don't think the alternatives have been tried in a serious way. Maybe one way to look at it is this: min wage says "work with a marginal product less than X is not worth having a person do; find another way or do without", and maybe in some sense we've decided we collectively prefer that somebody who might otherwise do X do nothing and receive social assistance payments. In particular I'm thinking about some of the truly awful jobs people in the developing world do, like melting down heavy metals in electronic scrap in their cooking pots (thus poisoning themselves), or the Untouchable Indian sewer cleaner I read about in National Geographic whose job it was to wade chest deep in faecal matter wearing nothing but a loin cloth. Perhaps we really are all better off pricing those jobs out of existence. It's a sort of blanket regulatory regime that nobody can evade with clever lawyers or corrupt/captured regulators.

And if the total benefit of X is high, but the marginal product of the labour to accomplish X is low (i.e. garbage collectors?), then isn't precluding human labour an incentive to some enterprising entrepreneur to find a way to build a cheap robot to do X? Then everyone will buy it and the clever business person will make a billion dollars (and thus become an evil plutocrat!).

Conversely, in a future where technology has obsoleted most humans, we might decide to reserve some work for humans. We already do that today in medicine. Computers are, and have been for several decades, better (in some case much better) at diagnostics than humans, but people in the city where I live (Edmonton) will still sit in the hospital emergency waiting rooms for 16 hours to see the inferior human. The computer isn't even offered as an alternative - probably because nobody would use it.

[Craig Burley]

"The computer isn't even offered as an alternative - probably because nobody would use it."

The computer isn't even offered because the humans have established a legal monopoly over the practice of medicine and anyone trying to offer the computer as an alternative would get prosecuted/sued out of existence. (As a lawyer, I know from professional monopolies!)

[himaginary]

FYI, it seems that the debate concerning minimum wage occurred a while ago, and I think this blogpost by Australian economist John Quiggin is a good summary of it.
And another Australian economist, Bill Mitchell of MMT, also supports minimum wage.

[Peter B.]

That EI increase is bad policy during a time of high unemployment. EI increase causes fewer jobs to be created which causes more people to be unemployed and collect more EI. Why not raise the GST to 6% instead.

[Robert McClelland]

That EI increase is bad policy during a time of high unemployment. EI increase causes fewer jobs to be created which causes more people to be unemployed and collect more EI.

Yes, that $7-$35 per year per employee is really going to upset the apple cart.

Why not raise the GST to 6% instead.

Because the rubes think its important that EI revenue (the change in the left pocket) be administered separately from general government revenue (the change in the right pocket). Therefore GST revenue cannot be used to balance the EI budget.

[Adam]

"If there was an alternative to minimum wages that achieved the same or better outcomes, wouldn't we expect to see the minimum wage being replaced by this other institution?"

Like how you'd expect governments to realize that, for example, the war on drugs and economic protectionism do nothing but harm their citizens?

It's not hard to come up with a list of policies that are completely, utterly, totally wrong and should be ended immediately. Our lists may differ, but expecting governments to stop doing something just because it's a bad idea seems rather bizarre.

[Robert McClelland]

I wonder if Nick can square what he wrote a couple of days ago with what's being said here.

Here's how non-economists think:

1. Why does the demand curve for apples slope down? "Because if the price of apples rises, people can't afford as many apples."

Obviously you need to substitute labour in place of apples.

[Nick Rowe]

Yep:

Here's how non-economists think:

1. Why does the demand curve for (low-skilled) labour slope down? "Because if the price of (low-skilled) labour rises, employers can't afford as many (low skilled) workers".

[Nick Rowe]

A hunch:

Suppose you were a non-economist who thought that the *only* reason demand curves sloped down was because of income effects. "If the price of X goes up, we won't be able to afford to buy as much X". But you were also a smart non-economist, who realised correctly that this explanation of the downward slope of the labour demand curve did not make sense at the aggregate level. Because if you raised minimum wages, minimum wage workers would see their incomes increased. So, in total we all *would* still be able to afford to buy the same quantity of labour. You would then conclude that the idea that an increase in the minimum wage would cause a reduction in employment was false.

[Adam P]

Nick, I think you've nailed it.

Not only would such a person conclude that the idea that an increase in the minimum wage would cause a reduction in employment was false but that person would believe that a rise in the minimum wage to $50/hour would, instead of reducing employment, cause inflation.

[Nick Rowe]

Yep. Because a small increase in the minimum wage could re-distribute income, but beyond a certain point would progressively just have to come out in inflation. I think that makes sense of how some minimum wage proponents might think. (Leaving aside the monopsony argument.)

[Robert McClelland]

Because a small increase in the minimum wage could re-distribute income, but beyond a certain point would progressively just have to come out in inflation.

That's what happens or would happen though. Increase the minimum wage by $1 and the cost of your Big Mac meal rises by only a small amount and doesn't result in a significant rise in inflation. Increase the minimum wage to $50 though and you're now paying a significantly higher amount for not only your burger and fries but almost everything else too.

[RSJ]

In my experience, most people don't really believe that wages are set by the marginal product of labor -- their own experience of says that wages are set by a lot of factors, particularly bargaining power, and from that point it's fairly plausible to argue that low income workers have proportionately less power than the top of the distribution, which leads you to believe that they are being paid too little, relative to their marginal product, whereas the top end is being paid too much, relative to their own marginal product. In a hierarchical organization with division of labor, it's very difficult to even determine the marginal product, and such a value may not even be defined. Just think of all the roles that workers have in an organization. Even a simple firm will contain product managers technical writers, interns, VPs, junior engineers, senior engineers, facilities management, etc. How are you supposed to define marginal product for any of these people?

That's the default "non-economist" view, and this is the basic rationale for using policy to put upward pressure on lower end wages and downward pressure on the highest wages. I don't think anyone believes that if you increased the minimum wage to some obscene amount -- such as $50/hr, that this wouldn't cause unemployment. But no one is advocating for that. Something simple -- such as keeping the minimum wage rising with labor productivity -- would already call for a roughly 30% minimum wage increase in the U.S., and this shouldn't increase unemployment at all.

[Robert McClelland]

Because if you raised minimum wages, minimum wage workers would see their incomes increased. So, in total we all *would* still be able to afford to buy the same quantity of labour.

Isn't that what you said in your post?

[Adam P]

RSJ:"In my experience, most people don't really believe that wages are set by the marginal product of labor -- their own experience of says that wages are set by a lot of factors, particularly bargaining power, and from that point it's fairly plausible to argue that low income workers have proportionately less power than the top of the distribution..."

two comments, first of all mainstream economics has plenty of models in which wages are higher than a workers marginal product.

Second, there is a good reason why there aren't any models predicting workers being paid less than their marginal product. The reason is that this would not be a situation that the *firms* are happy with.

The perfectly competitive market theory does *not* say that wages equal marginal product because workers refuse to work for less.

[Nick Rowe]

Robert: "Isn't that what you said in your post?"

Yes, I did say that (in effect). And it's (generally) correct (subject to possible minor qualifications we can ignore here).

The key point is this: economists (generally) do not (should not) explain the downward slope of the labour (or any other) demand curve by saying "If the price of labour increases, we won't be able to afford to buy as much". That's an "income effect" explanation, and it's (basically) wrong.

It's the substitution effect, and not the income effect, that economists (generally) do and should use to explain why demand curves slope down.

[RSJ]

Adam P,

"Second, there is a good reason why there aren't any models predicting workers being paid less than their marginal product. The reason is that this would not be a situation that the *firms* are happy with."

Huh? I'm talking about differences in wages across occupations, not industries. The *same* firm has both top management and lower paid employees. Why would "the firm" not be happy if the top management earned more than their marginal product, but the lower income employees earned less. Wouldn't the management have a greater say in deciding what the firm is and is not happy with?

What you've seen over the last 30 years in the U.S. is that 80% of the wages have not been keeping up with labor productivity whereas 20% of the wages have been outpacing labor productivity. Now that could be because management and finance is becoming much more productive, or it could be that they have managed to seize some of the wage shares of those with less power, leaving overall wage shares roughly the same. Do you disagree that this has been happening, or is it just that there are no models in which firms have heterogenous employees with different levels of power, all competing for a common pool of labor compensation?

[Adam P]

"I'm talking about differences in wages across occupations, not industries.."

So? Entirely irrelevant to the discussion. As is the rest of that paragraph.

As I've already stated there are plenty of reasons why firms may pay their employees more than their marginal product.

The reason firms wouln't be happy with a situation where a class of workers earned less than their marginal product is because this would imply the firm was failing to maximize their profits.

If the wage rate of a class of workers was less than their marginal product then firms would want to hire unlimited numbers of them, thus driving their wages up.

*In equiilibrium* these unskilled workers have no bargaining power, just as you suggested. In equilibrium the firm is indifferent to hiring the marginal worker or not. However, if the wage fell below their marginal product then they'd suddenly find that they have a very, very large amount of bargaining power as firms competed to hire them.

[RSJ]

"If the wage rate of a class of workers was less than their marginal product then firms would want to hire unlimited numbers of them, thus driving their wages up."

Why? The only proof you have of this assumes that the firm has only one type of worker. Unless you have some other proof?

If the wage rate of other employees was greater than their marginal product by an equal amount, then the firm would not want to hire any more labor -- the firm (i.e. the owners of the firm) would be satisfied that they are earning their required return. The owners of the firm don't care which employee is getting how much. They only care about their own return.

The management of the firm would be satisfied because it is earning wages in excess of its marginal product.

The other employees would accept the lower wages, with a reduction in labor supply. But the individual firm does not care about the reduced labor supply, it only cares about the equilibrium wage for each class of worker.

[Rick]

What I don't understand is the need for a Cartesian dualist framework where we have a distinct group of laborers and owners at odds. If laborers had a legislatively mandated stake in the company there would be no issue, raises would occur naturally as the company became more profitable.

There is a point when discussing the "economic" rationale for job losses wholly or in part due to minimum wage increases for low skilled laborers who were part of companies that doll out billions of dollars in dividend payouts and executive bonuses amounts to a disservice for the population at large. Unless that rationale is accompanied by, that's why we need strict regulations or creative taxation to prevent that kind of outsourcing.

Minimum wage increases weren't designed to inconvenience small business, and it is unfortunate that in practice they do, but they are the only way to influence large businesses that can't get away from having huge numbers of domestic employees.

And as an aside, Mike... You don't really earn your money in the sense that you have any business outsourcing web work to India (and to boast about it???), where would you be without the regulatory subsidy you got as a Canadian Student/Professor???? So easy we to forget where you came from....

[RSJ]

What I have in my mind is a series of nested subcontracts: A_0 hires A_1, A_1 hires A_2, etc.

In the coarsest analogy, A_0 would be the owners of capital, A_1 would be CEOs and "top" management, and A_2 would be everyone else.

But you can keep extending the nest.

In this case, what prevents A_1 from skimming from A_2, even while it does not skim from A_0?

As long as shareholders need management to run the firm -- i.e. they cannot disintermediate, then they will not be able to take advantage of the reduced wages paid to the A_2 in order to increase the demand for A_2. But A_1 will not increase the demand for A_2 as this cuts into their labor rents (as the sum of payments A_1 + A_2 is fixed at the overall market price).

I'm not sure if this type of nested subcontracting is well covered in theory somewhere, but it matches experience. You've had an enormous increase in compensation of CEOs and other top earners even as overall factor payments have not moved a whole lot. And this movement coincided with a decrease in labor union power.

[Blikktheterrible]

"I still have to drag the phantom job losses canard (ie. without the minimum wage increase we're making a random guess that there would have been more jobs created) out of the closet."-Robert McClelland

So, you're arguing that we have no way of demonstrating any link between employment levels and the minimum wage? Is that a correct characterization of your argument? Ok, I'll assume that it is and that you're correct. Under those assumptions, we would be left with the orthodox theory that raising the minimum wage will decrease employment, which is based on reasoning from mainstream economic assumptions, and no way of seeing if that theory matches reality. To put it another way, expert intuition is all we have to go on. In that case, I would argue that we should assume that expert intuition is correct. There's not much reason to accept it, but there's even less reason to reject it.

Do you disagree with any of this?

"The people who rail against minimum wage don't want a replacement. According to them, life was just grand for everyone when no minimum wage existed so we should just go back to having none at all and let the free market dictate how many cents per day the working scum should receive."--Robert McClelland

I don't think that's quite. It's probably true that many people who oppose (raising) the minimum wage would not want it replaced with anything, but I don't think that's true of the people on this discussion board. It's certainly not true for me. I'm marginally against the minimum wage and I would be very happy to see it replaced by another system, even if that system were quite expensive.

"If there was an alternative to minimum wages that achieved the same or better outcomes, wouldn't we expect to see the minimum wage being replaced by this other institution?"-Frances Woolley

I think the key factor is public perception. No one feels harmed by the minimum wage, but many people feel supported by it. I don't think that perception is accurate, though.

[RSJ]

"To put it another way, expert intuition is all we have to go on. In that case, I would argue that we should assume that expert intuition is correct. There's not much reason to accept it, but there's even less reason to reject it."

No, what we have to go on is politics. The expert intuition should override the political conclusion only to the degree that the experts have a robust theory that is accepted by the body politic as being impartial and accurate.

If not, then the voters and politicians decide based on their own whims, and they should weigh the influence of current economic conventional wisdom by the overall track record of the discipline.

Economics isn't like engineering. Voters/politicians are happy to outsource the design of bridges to experts because we have a scientific body of laws that are sufficiently accurate to allow the building of bridges that don't fall down. This is generally accepted by everyone, and so you don't see politicians deciding the best way to design a bridge.

If economics ever reaches a similar level of accuracy, *then* we can offload discussions of the minimum wage to experts, but not until then.

[Adam P]

RSJ: "The only proof you have of this assumes that the firm has only one type of worker"

No, another utterly false statement. It's true worker by worker even if every single one has a different level of skill.

"If economics ever reaches a similar level of accuracy..."

How do *you* know it hasn't?

[Nick Rowe]

Rick: you have a very different view from orthodox economics on how firms ought to set wages and employment. It's a view I hear often. Roughly, it goes something like this: "Employers should never cut wages and/or employment unless they can't afford not to. Employers should increase wages and/or employment when they are making high profits."

The orthodox economist's view goes roughly like this: "Employers should cut wages when there is an excess supply of labour, and raise wages when there is excess demand. Employers should increase employment when an increase in employment would increase profits, and cut employment when a decrease in employment would increase profits."

(Both are simplifications, of course). Those are very different perspectives.

[Mike Moffatt]

"It's the substitution effect, and not the income effect, that economists (generally) do and should use to explain why demand curves slope down."

Thanks Nick! Note the initial post that prompted all this discussion was about substitutes to low-skill labour (technology, outsource, etc.)

[Nick Rowe]

Mike, yes I noticed that. You specifically cited substitution effects towards "possible alternatives".

[Mike Moffatt]

Just thought I should remind our audience. :)

[RSJ]

Adam:

OK, "utterly false" is not an argument, but I will venture to say that you are assuming that each factor of production adjusts separately, so that if there are 10 factors with MPi somewhat higher for some and somewhat lower for others, then all of these will shift to bring them each separately into balance, rather than having them shift to bring the sum into balance. The latter exhibits more degrees of freedom that can be influenced by bargaining.

Whereas your model where you hire a second programmer and then have him share the same computer because labor is the "variable" factor of production whereas capital is the "fixed" factor, so that if capital costs too much and labor costs too little, firms will just hire more labor and keep the capital fixed.

That's highly speculative, and I don't firms operate this way. Not in all cases, at least.

Moreover, when it comes to top management -- e.g. CEOs, etc., then there isn't much freedom for a firm to reduce it's utilization of this factor -- that's part of what it means to have bargaining power. If top management is earning labor rents, then there will be either unemployment of the other factors of production, or the other factors of production will be employed, but they will earn less than their marginal products. In order to say with confidence that only the second situation will occur, you need a separate adjustment result and I don't think you can plausible argue that there is such a result. You can, of course, *assert* the result, and loudly if you want.

In terms of how do I know, then this is the whole point. The burden of proof is on the economists, not on the public. They need to convincingly prove that they can predict what will happen to the economy if a certain policy is enacted. A good start would not be getting blindsided by every crisis, or being able to do better than a toin coss at predicting the macro variables. If they can't do that -- and perhaps this is impossible -- then they have no place at the policy table.

[Nick Rowe]

The firms chooses L to maximise profit = P.F(L) - W.L where L and W are vectors of 10 factors and their 10 wage rates, and F(L) is the production function.

You get 10 first order conditions of the form:

dprofit/dL1 = P.dF(L)/dL1 - W1 = 0

Price times Marginal Product = Wage