Some economic policy debates are about jobs. Most aren't about jobs, or shouldn't be. But far too many of them end up being about jobs, even though the policies have nothing to do with jobs.
Frances once explained the advantages of the HST to me. (Note for non-Canadians: the Harmonised Sales Tax replaces a provincial sales tax with a VAT collected in conjunction with the federal VAT). She talked for a few minutes about collection costs and the deadweight loss triangles of distorting taxes. She never mentioned jobs. I didn't ask her whether it would create jobs. If I had asked her whether it would create jobs, I think she would have given me a very quizzical look, then shrugged her shoulders. The HST is not about jobs.
So I was very surprised to read Iglika Ivanova's critique of Jack Mintz's argument that the HST would create lots of jobs. It's not the critique itself that surprised me, but that the debate existed. "My God! People are actually arguing in favour or against the HST based on how many jobs it will create or destroy?!" I was stunned, though perhaps I shouldn't have been. The debate over the Canada-US Free Trade Agreement was mostly about jobs too, though the benefits of free trade have almost nothing to do with creating jobs.
Economies are a general equilibrium system. Everything depends on everything else. If you change any one thing, everything else will almost certainly change too. And that includes HST and jobs.
Here's my guess on how the HST will change employment. I think the HST will probably make us a little bit better off. When people are better off, they usually choose to work a little bit less. I think the HST will probably cause a very small reduction in employment.
My guess about the effect of HST on jobs could easily be wrong. If the HST is a more efficient tax than the one it replaces, real wages will probably rise, as part of the general improvement in our welfare. And maybe the substitution effect of the increase in real wages will draw people out of leisure, or home production, into working more. And that substitution effect might be bigger than the income effect, so the net result is an increase in employment. Maybe. It all depends.
But the key point is that whether my guess about the effect of HST on employment is right or wrong has nothing to do with whether I would support or oppose the HST. If it makes us better off by reducing collection costs and tax distortions, then it's a good tax. It doesn't matter whether people choose to respond to those benefits by working more or working less.
Now it's true that some taxes, like direct taxes on labour income, will tend to distort the labour market primarily, and make us worse off precisely because they create a wedge between what the employer pays and what the employee gets. (I'm setting aside the benefits we may get from the things that tax revenue is spent on, of course.) And the deadweight costs of those taxes are due to the reduction in employment they cause, if either the labour supply or demand curve is not perfectly inelastic. But even in that case, the costs of the tax are not measured by the number of jobs lost; they are measured by the loss in welfare created by the cost of the loss in output minus the benefits of working less. At the margin, we work in order to produce output which we can consume or invest. At the margin, work is a cost, output is a benefit. (If it weren't, and we were all prepared to work just as many hours for free, then taxes on labour income wouldn't reduce employment, and wouldn't have any deadweight costs.)
Employment is not a measure of well-being. If we could get the same output while working fewer hours, that's a good thing. We can spend those extra hours doing something else we would prefer to be doing.
So why do so many debates end up as debates over jobs created or destroyed, even when the policy in question has nothing to do with jobs? That's what I can't figure out.
I once saw a poster advertising a Carleton seminar on "Women in society and X". I can't remember what "X" was. All I remember is that X was an important issue for many reasons, but none had anything to do with the relative roles of men and women in society.
I once shared an apartment with two communists. One was a Glasgow Rangers supporter, the other a Glasgow Celtic supporter (football teams). I remember them arguing whether a Rangers or a Celtic victory would be best for the revolution.
Novelist Mordecai Richler once parodied the sort of debate I am talking about. Something like: "Headline: 'Tiger escapes from zoo'; response: 'Is that good for the Jews or bad for the Jews?'
I call these debates "more studies in applied orthogonality".
I think we sometimes get fixated on a certain dimension, and analyse all policy questions in terms of how they look on that dimension, even when they matter on some other dimension that is orthogonal to the first. Involuntary unemployment is a problem. It's a dimension that matters. But reducing involuntary unemployment is not always the same as increasing employment. And some policies are good policies for reasons that have nothing to do with reducing involuntary unemployment, or even increasing employment. I can't think of any reason why the HST should have any effect in either direction on involuntary unemployment.
Short-run macro policy is about jobs. So is the minimum wage. So are unions.
The HST is not about jobs. Nor is free trade. Nor is environmental policy like carbon taxes. Most economic policies aren't about jobs.
Nick, I was going to write an extensive response but I'm heading off to UBC to tour the campus and hit the beach (no, not that beach, my son is too conservative).
Yes, I would give a response in terms of deadweight loss triangles. But how do you explain intuitively to an undergraduate what that deadweightloss triangle actually respresents?
A deadweight loss means resources aren't being used as efficiently as possible. So they're not employed as efficiently as possible. They're un- or under-employed. You can see how that can easily slide into a discussion of jobs.
Posted by: Frances Woolley | July 09, 2010 at 02:16 PM
I think discussions about 'jobs' make a lot more sense if you think of "jobs created" as "increase in the demand for labor" and "jobs lost" as "decrease in the demand for labor". I think 'jobs' is really a lay term for labor demand. Of course in many cases the demand for labor is still orthogonal to the issue at hand.
Posted by: jsalvati | July 09, 2010 at 02:38 PM
Frances: agreed, but with the HST it sounded more like the benefits were getting the right number of people working in the right places, rather than getting the right number of people working in total.
jsalvati: I think you are on to something there. I'm not sure if I've got it right, but I think that what Jack Mintz has done is equivalent to taking a GE model, imposing a horizontal labour supply curve, then asking how much employment would increase in that hypothetical scenario. That's equivalent, I think, to asking how much the labour demand curve shifts right. But the outcome to that thought experiment bears very little resemblance to the increase in employment predicted by the model with a true labour supply curve (obviously), or (less obviously) with any measure of the benefits of the HST.
Suppose for example that the labour demand curve were also very elastic. In that case, a tiny subsidy to employment (or cut in taxes on labour) would shift the labour demand curve vertically up a tiny amount and yet cause a massive increase in employment. In the limit, with a perfectly elastic labour demand curve, "jobs created" by a tiny labour subsidy or cut in labour taxes would become infinite.
And the poor common people, on hearing about this, would think it's the best economic policy ever imaginable!
The right thing to do, in my humble opinion, would be to calculate the effect on utility of the representative agent, and then calculate an amount of manna from heaven that would give the same increase in utility, take the nominal value of that manna at current prices, and then say "The HST will make the average person $10 better off". Or, "The HST is equivalent to a 0.1% increase in people's incomes".
By why "Jobs!" as the metric? And why that way of measuring the jobs-equivalent?
Posted by: Nick Rowe | July 09, 2010 at 05:13 PM
"By why "Jobs!" as the metric?"
The answer is that 'jobs' is the only metric that most voters seem to understand. Hence, politicians--left, right, and in between--want everything quantified in terms of jobs.
So, if you are an economist wanting to make an impact on a political debate you can:
a) talk about the triangles and long-run equilibria and be mostly ignored.
b) translate things into some crude measure like 'jobs' and not be ignored.
If you are doing (a) and your opponents are doing (b), you will likely have a nice moral victory to savour, but your advocacy will not have much political success--nor ultimate impact.
Different people have different thresholds about how much (b) they are willing to mix with their (a).
p.s. Frances--you picked a nice day to tour campus. Let's just hope your son assumes the weather is like this 365 days a year . . .
Posted by: Kevin Milligan | July 09, 2010 at 08:55 PM
Kevin: Oh God. You mean that even in those cases where we actually do know something about the subject, like the HST, we don't say it. We talk about total bullshit instead, like Jobs! Race to the bottom.
How about: "every hour we delay implementing the HST, a kitten dies."?
I could probably make a CGE argument for that, about as plausibly as the jobs argument, if you gave me a grad student with a computer, plus a wad of cash.
Posted by: Nick Rowe | July 09, 2010 at 09:40 PM
How about: "every hour we delay implementing the HST, a kitten dies."?
You know, I've been looking for a catchy slogan for the census campaign. Mind if I steal this one?
Posted by: Stephen Gordon | July 09, 2010 at 09:59 PM
Why is the jobs metric so difficult to understand? A job and the pay that results from it constitutes 90% of the interaction with the economy for most people.
RRSP's, Mutual Funds? Hardly. They're intermediated by salesmen.
Jobs = pay. If you can't relate it to job compensation and money, it won't get through.
Posted by: Determinant | July 09, 2010 at 10:21 PM
I had my dad read the post by frances and it calmed him down, so don't dumb it down. I remember a commenter on coyne's blog asking for an honest opinion a year ago, saying he couldn't find info... a lot of people do not have very good sources of information. Not to say we've had bad journalists, but the internet also helps with asynchronous search requests.
Posted by: edeast | July 09, 2010 at 10:38 PM
I think discussions about 'jobs' make a lot more sense if you think of "jobs created" as "increase in the demand for labor" and "jobs lost" as "decrease in the demand for labor". I think 'jobs' is really a lay term for labor demand. Of course in many cases the demand for labor is still orthogonal to the issue at hand.
As a mostly lay (first year economics student) person who uses mostly lay terms, who has been in many an HST debate since BC is the hotbed of HST debate, I would have to concur that "jobs" is exactly that; no one would argue that a policy that increases employment among people who don't necessarily want or need to work is necessarily a good thing, but everyone intuitively would argue and does believe that to create a situation where it is easier for people who want to work to obtain work, or to create a situation where the ones who are working get paid a little bit more, is a good thing. This, in my limited understanding, is pretty much synonymous with an increase in the demand for labor, and when talking to the average lay person this is exactly what "jobs" means.
It would be difficult (read: impossible) to convince the average layperson to care about deadweight loss. It's just not an intuitive abstraction without sufficient formal instruction, and it's difficult for pretty much anyone to accept that 1) deadweight loss matters and 2) minimization of deadweight loss will impact theirs or anyone's life in a meaningful way. To the average layperson, the fact that haircuts and restaurant meals (and a whole host of other things) cost 7% (in BC) more than they did on June 30th is of much more concrete, immediate consequence than the minimization of deadweight loss. Any attempt to argue that the minimization of deadweight loss offsets that perceived increased cost of living will be met with scorn and accusations of either being blinded by science or some kind of corporate/Liberal shill or just a troll.
"Jobs", on the other hand, is a concrete concept, and a powerful one at that. Anyone can appreciate the difference between a person who wants to work having a job, and a person who wants to work not having a job, and can weigh an increase in "jobs" (jobs in this context being defined as above) against an immediate 7% increase in taxi fare. And, correct me if I'm wrong, but isn't a side effect of HST's deadweight loss reduction that investment is increased, which leads to an increase in demand for labor? And, if "Jobs" means demand for labor, and increased demand for labor is both an effect of HST and judged, generally, to be a good thing, then wouldn't making a "jobs"-based argument be a good way to go?
Posted by: Colin | July 09, 2010 at 10:42 PM
Determinant: "Why is the jobs metric so difficult to understand?"
Oh it's easy to understand all right. But to say that the HST is good *because* it creates jobs is *false*. It might or might not increase employment. But that's not what makes it good.
Colin: "And, correct me if I'm wrong, but isn't a side effect of HST's deadweight loss reduction that investment is increased, which leads to an increase in demand for labor?"
Like any change to a less distorting tax system, the HST should lead to an increase in investment in some areas, and a decrease in others. I'm not sure what it will do to total investment. And an increase in investment (and in the stock of capital) may or may not lead to an increase in the demand for labour. It usually has done historically, but it doesn't have to. In any case, I can imagine the possibility that a change to a less distorting tax system might reduce output of labour intensive goods, and increase demand for capital intensive goods (or land intensive goods). In which case the demand for labour might fall, and real wages and employment might fall. Unlikely, but possible. And again it wouldn't affect the desirability of the tax change.
If we said that the new tax would increase people's real incomes by x% on average, that is something people should understand. And it would be at least close to being intellectually honest. (Strictly we should say that it would increase their economic well-being by an amount *equivalent to* an x% increase in real income. But hey, it's close enough.)
Posted by: Nick Rowe | July 09, 2010 at 11:47 PM
Hi Nick,
I'm not sure if your lament was aimed at me or the world in general.
My claim was positive, not normative.
You asked why 'jobs' is the metric. My answer is because 'jobs' is an argument that voters understand, so that is an argument that politicians use. I might wish otherwise, but alas my wish is not reality.
Posted by: Kevin milligan | July 10, 2010 at 01:02 AM
If your question is why *economists* are making these arguments, then it comes back to the tradeoff between making subtle and serious arguments and being ignored, vs making more populist facetious arguments and influencing the debate.
You and I may prefer to keep things more rigorous, but then we have to accept that we will likely be ignored in a debate that is quite literally played out in the streets, here in BC anyway. What argument will sway the 'stop HST' signature gatherers who have spent much of the last 6 months huddled rain or shine in Robson Square a few blocks from where I type this. It's not me you have to convince--it's the angry person whipped up by Bill Vander Zalm's rhetoric.
That isn't to say that economists should all abandon rigor and start threatening the kittens. But neither should we fully retreat to our seminar rooms and leave the debate on key matters of economic policy to the non-economists. I wish I had a good answer to that conundrum.
One solution is to have a blog where smart economsists can point out when policy debates are way off track; trying to communicate as clearly as possible the subtle and serious analysis. (Thank you, WCI writers!) Maybe that will influence a few people anyway.
But until we figure out how to communicate things like Harberger triangles in 10 second sound bites, I think we're going to be stuck with a world where politicized policy debates (like the debate on HST) lack the rigor of the seminar room--or even the Econ 101 lecture hall.
Posted by: Kevin milligan | July 10, 2010 at 02:01 AM
diversion. These are old but i like them.
0:45 of this
1:45 of this
Posted by: edeast | July 10, 2010 at 03:01 AM
After John Turner made the world's silliest pleonasm - "Jobs - for people!" - a key slogan of his election campaign (in 1984? 1988?), I figured out that 'jobs' just meant 'something good.'
Posted by: GA | July 10, 2010 at 04:49 AM
Kevin: I was just howling at the moon. You were the messenger bringing bad news.
But I still think my "It will [(sotto voce) be the equivalent of an] increase incomes by x%" is better than "It will increase jobs by x%". More correct, and as easy to understand.
Posted by: Nick Rowe | July 10, 2010 at 08:02 AM
Some day, when the goods of the world are entirely produced by robots it ought to be obvious that social policy is about equal distribution of land and monopoly rents. But even then, some socialist will be yelling about jobs like it's good for "the people" to be digging ditches and filling them back up again. And the rentiers will love it. Since "jobs" are arbitrarily cheap it will be a great way to soothe the rabble.
From the median on down, there hasn't been a raise in 40 years. But lots of mcjobs. It's not pretty where we are heading.
Stuff good, work bad. (And monopoly rents are a communal good). The sooner we all start getting it, the faster we can get on the right course.
Posted by: K | July 10, 2010 at 08:05 AM
Hi Nick,
I agree that x% more income over the next 10 years is a better metric. Why don't we hear so much of that? Ask the next political consultant you run into. Myself, I'm much better with the Harberger triangles than the focus groups.
But if you want my WAG, then I would say that the emotional impact of a new job (or especially the loss of one)>> the emotional impact of an x% increase in income over the next 10 years.
Just think of opposition politicians (A) and (B) on the hustings:
A) Your policies have cost us a 4% increase in income over the next decade. For shame!
or
B) Your policies have cost 10,000 families their livelihoods! Tiny Tim shall have no Christmas goose. For shame!
Posted by: Kevin Milligan | July 10, 2010 at 02:40 PM
Nick - I think your beef might actually be with people who do partial rather than general equilibrium analysis. Take the construction industry in isolation. HST increases the 'tax wedge', the difference between the amount consumers pay for a new house and the amount producers receive. So you'd expect either fewer consumers demanding houses or fewer producers being willing to produce houses - either way, less houses get built, so there's fewer jobs.
But now think about the economy as a whole. The old retail sales tax (RST) taxed some industries relatively heavily (ones that produced taxed outputs and also used taxed inputs) and other industries (services) relatively lightly. So in the long run one would expect some prices to go down and some prices to go up - and production (hence employment and jobs) to increase in industries that are now in a better tax position (exports) and decrease in those that are in a worse tax position.
I think what you're trying to say in your post is that 'hello, it's not that suddenly resources are going from being unemployed to being employed, it's that they're being used more efficiently - to produce things people actually want, or to produce things with fewer inputs, or to produce more stuff.'
But general equilibrium analysis is much much harder to do than partial equilibrium analysis.
Posted by: Frances Woolley | July 11, 2010 at 03:54 PM
"But far too many of them end up being about jobs, even though the policies have nothing to do with jobs. . . .
"The debate over the Canada-US Free Trade Agreement was mostly about jobs too, though the benefits of free trade have almost nothing to do with creating jobs."
I am not arguing against free trade. However, according to Bill Mitchell, the arguments for free trade assume full employment. (My take is that they assume that employment is irrelevant.) If he is wrong, please say why.
Now, among Canada, the U. S., and Mexico, the poor neighbor is Mexico. Labor costs are generally lower there. (Without the free movement of labor, free trade is not going to raise those costs anytime soon, right?) So say that capitalists tend to relocate from the U. S. or Canada to Mexico. If that movement of capital increases unemployment in the U. S. or Canada (because those jobs are not protected by restraint of trade), then the assumptions of the free trade arguments are violated, right? So jobs are relevant, no?
Posted by: Min | July 11, 2010 at 05:52 PM
The way we usually tell the story is that the reduction in labour demand would generate lower wages - the famous Stolper-Samuelson theorem.
Posted by: Stephen Gordon | July 11, 2010 at 06:21 PM
"Employment is not a measure of well-being."
Right. :)
"If we could get the same output while working fewer hours, that's a good thing."
Ceteris paribus. :)
"We can spend those extra hours doing something else we would prefer to be doing."
Talk to Sandwichman. ;)
But who is "we", White man? In the developed world during the last generation, increases in productivity may have increased aggregate per capita income, but have not increased either median incomes or leisure appreciably. OTOH, unemployment has been high in Europe for some time, and the last two U. S. recessions have produced persistent high unemployment. Those extra hours are not being spent doing something else that we would prefer.
With the dawn of the Industrial Revolution, Benjamin Franklin and others looked to a future where labor-saving devices would produce general prosperity and leisure. It has not worked out that way, for a number of reasons. While employment is not a measure of well-being, given the givens of society, unemployment is an index of the health of an economy. If we lived in a relative utopia, things might be different. But without a job, most people cannot share in the general well-being of society.
For the general public then, questions of economic policy will normally raise questions of employment. In the complex system that is a modern economy, policy effects are not always obvious to the public. If the policy is not related to employment, most of the time people need to be reassured of that.
Posted by: Min | July 11, 2010 at 06:35 PM
@Min: We don't live in a utopia, but labour-saving devices have most certainly produced general prosperity and leisure. To take two everyday examples: because of treatment technolgy, plumbing and distribution systems, and the material and construction technologies that make them all possible and affordable, we don't have to haul a day's worth of water from the well anymore, and we don't get sick from drinking it. And for a small capital investment, and a willingness to sit through ads, a night of television is *way* cheaper than a night at the theatre or a sports game. We can find literally thousands more examples in all fields of human endeavour.
I even challenge your point in that paragraph: they certainly don't get a large share, but the unemployed most certainly do benefit from the general prosperity -- as bad as it gets for them, it would be worse if society was poorer too.
Posted by: Two Hats | July 12, 2010 at 11:09 AM
@Nick:
This is somewhat off-topic, but I don't know what other thread to ask you on. ;)
You have said before that labour is powerful. If that is the case, why does it seem that the majority agree with Min's statement on leisure? Even speaking for myself, and I am not poor, I find it very difficult to turn down overtime. If labour was that powerful, wouldn't we have enough money to be comfortable without scrabbling for extra shifts? Wouldn't we have negotiated the number of hours we work down to something more manageable, like 30h/week, but have insisted on a maintained salary? It seems to me that Capital is really much more powerful, and that one of the reasons income inequality has been growing so greatly for the last 30 years.
Posted by: the_iron_troll | July 12, 2010 at 03:02 PM
Kevin: Unfortunately, you may be right. Think of Tiny Tim, whose Dad works collecting the PST!
Frances: Agreed. But Jack Mintz was doing general equilibrium analysis (or general disequilibrium, because he held prices fixed!). And what I'm also trying to say is that your short comment just there, for example, tells us far more about the HST than any such "Jobs!" analysis.
Min: Bill Mitchell is essentially correct. To put it more precisely, arguments in favour of free trade generally assume that there is no *deficient aggregate demand* unemployment. And that's the right assumption to make, because:
1. On average there usually isn't any deficient-AD unemployment
2. Because if there were deficient-AD unemployment, or if free trade might tend to make it worse, the obvious solution is not to avoid free trade, but to use monetary and/or fiscal policies to increase AD and eliminate it, regardless of what you are doing with trade.
So Bill is correct, but his point (if he's talking about deficient-AD unemployment) is beside the point.
There are other reasons for unemployment besides deficient AD. One of my colleagues spends a lot of his research trying to come up with models where free trade is not first-best when you have unemployment due to various other causes. My own judgment is that free trade, by making labour markets more competitive, will tend to reduce unemployment caused by lack of competition. But I think that effect will likely be minor.
"Now, among Canada, the U. S., and Mexico, the poor neighbor is Mexico. Labor costs are generally lower there. (Without the free movement of labor, free trade is not going to raise those costs anytime soon, right?) So say that capitalists tend to relocate from the U. S. or Canada to Mexico. If that movement of capital increases unemployment in the U. S. or Canada (because those jobs are not protected by restraint of trade), then the assumptions of the free trade arguments are violated, right? So jobs are relevant, no?"
No, basically.
First, even without free flows of labour and capital, free trade can sometimes (under some conditions) cause factor-price equalisation, so real wages equalise.
What happens to Canadian and US wages when we open up for free trade with Mexico may depend on why Mexican wages are lower initially. You may get equalisation somewhere in between, or you may not. If Mexican wages are lower because of worse technology, for example, and if free trade promotes the spread of better technology, then Mexican wages will tend to rise, rather than US and Canadian falling. They adopt our technology; we don't adopt theirs.
From my very cursory reading, the "free trade is what worsened wages" hypothesis doesn't fit the facts. Relative output prices didn't change the way that hypothesis predicted they would.
But look, in some cases free trade will change the demand for labour (and different types of labour differently) in either direction. But that means it can affect wages and the distribution of income, as Stephen says. It doesn't mean it's about "Jobs!"
"With the dawn of the Industrial Revolution, Benjamin Franklin and others looked to a future where labor-saving devices would produce general prosperity and leisure. It has not worked out that way,...."
Yes it has! Though most people have chosen to consume a lot more indoor toilets, cell-phones and other gizmos and only a little bit more leisure. Which is OK, if it's their choice. Ooops! Two Hats has already made this point better than me.
But you really need to distinguish: unemployment, employment, real wages.
Posted by: Nick Rowe | July 12, 2010 at 03:38 PM
The iron troll:
Assume initially one side of a market gets more power. If sellers get more power, they force price up. As price rises, this creates an excess supply of the good, which reduces sellers' power and increases buyers' power. Which makes it harder for sellers to raise prices. In equilibrium prices are higher, but power is equalised again, so price neither rises nor falls from its new higher equilibrium.
But there's a whole different question you are mixing in here: what do you want: leisure, or money? Looked at from our grandparents, we are ridiculously rich. Why don't we all work half-time, or take every other year off, or save half our income and retire at 50? That's got nothing to do with power. It's about preferences. Hippies say we're nuts. Maybe hippies are right. Save half your income, then quit!
Posted by: Nick Rowe | July 12, 2010 at 03:47 PM
I may just retire at 50, or maybe 60 if I can get some 30h weeks. I'm not nuts.
Okay, I think I understand. I don't know exactly how to fit unions into this equation, though. Do you think they're totally unnecessary, then, and labour would create a satisfactory equilibrium level without them? Are Wal-Mart employees paid what they're worth? I'm rather concerned with the welfare economics of the situation - I would like to increase efficiency, get rid of that nasty dead-weight loss, but I think that welfare is more important.
Posted by: the_iron_troll | July 12, 2010 at 04:25 PM