Why New Keynesian macroeconomists are against labour unions

Like all New Keynesians, I believe two things that are apparently contradictory:

1. We each get paid too much; 2. Because of 1, we all get paid too little.

It only makes sense if you understand fallacies of composition. What is true of each part isn't necessarily true of the whole.

It's a very Hobbesian perspective. It's not a war of Us against Them; it's a war of All against All. Those who can only think in an Us against Them perspective will have the hardest time grasping it. You have to think in a Hobbesian perspective. Or representative agent model, as we call it nowadays. The distinction between the individual and the collective good is what's at issue, not the distinction between the good of one group of individuals and another group of individuals.

I have been thinking like this since I was 18, but I only understood it vaguely back then. I understood it better when I was 30, and started building macro models with monopolistic competition.

Each individual faces a downward-sloping demand curve. Taking other individuals' prices as given, the higher I set my price the lower the quantity I can sell. That's because we are not all identical. Nobody is a perfect substitute for anybody else. So we are not in perfect competition. So each individual faces a trade-off between price and quantity.

Microeconomists are with me so far; here's where I'm in danger of losing them.

But the demand curve facing all of us is horizontal. There is no trade-off between relative price and quantity at the aggregate level. The price of everything relative to everything must be one. And because the trade-off facing each individual is different from the trade-off facing all of us, each individual will choose a point on the trade-off that is bad for all individuals.

OK, let's go back and round up the stray microeconomists (and probably a few macroeconomists as well).

Every microeconomist knows that it's relative prices that should be on the vertical axis of demand curves. Suppose one individual wanted to sell more. Unless he is a perfect substitute for other individuals, he will need to lower his price relative to other individuals' prices. But if all individuals want to sell more, and all lower their prices, the average relative price does not change. If we aggregate up over individuals' downward-sloping demand curves, we get an Aggregate Demand curve, with the average relative price on the vertical axis, that is horizontal at one.

And now I've lost some macroeconomists too. Because I've drawn an AD curve with relative price on the axis, rather than the nominal price level. But don't worry. Remember that the AD curve is an equilibrium locus. It's a set of points at which output equals output demanded. We can ask what happens to the nominal price level as we move along that set of points, and that's the normal AD curve. But we can also ask what happens to relative prices as we move along that set of points, and that's the AD curve I'm talking about here.

When we talk about being paid too much, we should always understand this in real terms. What matters is not how many dollars we get paid; it's how much we can buy with those dollars we get paid. It's the relative price, not the nominal price that matters.

Because each individual faces a downward-sloping demand curve, but all individuals together face a horizontal aggregate demand curve, each individual will choose a price that is too high and a quantity that is too low. Because one individual's selling price is another individual's buying price, all would be better off if all cut prices and increased quantity. Prisoner's dilemma. The attempt by each individual to raise his own real income above the competitive equilibrium results in a lower real income for all.

If all prices are sticky, then an expansionary monetary policy, which increases aggregate demand, increases output and makes everyone better off. That's why booms are good, because it brings the economy closer to the competitive equilibrium. And why recessions are bad, because they take the economy further away from the competitive equilibrium.

And cartels, like labour unions, just make the problem worse. Because by joining together with similar sellers into a group, the demand curve facing the group is  steeper than the demand curve facing the individual, since members of the group no longer compete against each other for buyers. So there is an even bigger difference between the downward-sloping trade-off facing the group of sellers and the horizontal trade-off facing us all.

Unions are bad for the very same reason that recessions are bad.

All New Keynesian macroeconomists have understood the above for the last 20 years. Which is why all New Keynesian macroeconomists are fundamentally opposed to cartels, labour unions, minimum wage laws, etc.. OK. It's why they should be opposed to such things.

Comments

[Lord]

And monopolistic competitors too, right? They are also a source of lumps in the stream, especially when they become big enough to be true monopolies.

[RSJ]

Who is "we", white man? :)

The rich are just as overpaid as much as the middle class? Are capital returns too high or is it just that labor costs are too high? They are both equally too high? All factor inputs equally cost too much? And there is no gain available from equalizing the amount by which people are overpaid?

Take a gander at the BLS paper "The compensation productivity gap: A visual essay".

What bothers me is that you are taking a highly stylized model that ignores dynamics, investment, heterogeneity, finance, the endogeneity of productivity, and concluding that you "know" that "people" are overpaid.

All you know is that if the hypothesis hold, then the conclusions hold.

"If A, then B".

But in this post, you forgot A, and then determined that you know B applies in our world, where A is manifestly false. And if your post is to be believed, then sometime in your 30s you decided to disassociate hypothesis from conclusion and just started making blanket statements about the world rather than making "if, then" statements:

"All New Keynesian macroeconomists have understood the above for the last 20 years. Which is why all New Keynesian macroeconomists are fundamentally opposed to cartels, labour unions, minimum wage laws, etc.."

There wasn't a single "if" anywhere in there.

No conditions under which these conclusions held, because the conditions are so absurd that no one would believe the conclusions as being remotely applicable to our economy.

But given the overwhelming data that a majority of the people are underpaid, one would want to go back to his hypothesis and try to change them to see if he could create a model that could predict the compensation gap, as well as recurring asset bubbles.

I'm sure that many are doing that -- I'm not slamming the profession (or you).

I'm slamming this particular post, in which you know, despite all the overwhelming evidence of stagnant wages, that people are being overpaid. And more over, you assert, that all other economists -- even the ones who prefer to stick to "if, then" statements -- must share your policy position vis-a-vis our actual economy.

You are sticking to your guns, to what you have known for 20 years. FYI, 30 years ago this compensation-productivity gap took off. Perhaps 20 years ago you were right in describing the world in the post-war era to about 1980. Perhaps at that time, unions were too strong, or the fantastical hypothesis that you rely on were, just by accident, not too harmful to the applicability of the conclusions reached.

But that's not the economy that we live in today.

[Patrick]

Can't really argue against the logic, as far as it goes.

I'd second what Lord said and add that the C-suite clique is a de facto union if ever there was one, and far more destructive than anything a group of teachers.

Given the experience in the US over the last 20 or 30 years, I have to wonder if it's really so simple: bust the unions and we're all better off? Really? And in practice it seems to me the focus (and vitriol) is always on some group of working shlubs - teachers for example - whose wages collectively represent a tiny, tiny, tiny fraction of the rents extracted by the likes of Jamie Dimon or the high rollers at AIG FP who damn near blew-up the world economy. If we're going to run around distortion busting, it seems to me that there are bigger fish to fry than unions of teachers and gov't clerks whose lifetime earnings are collectively probably less than the bonuses at GS in a single year. Of all the distortions in the economy, at present unions seems to be low on the list of major problems.

[Determinant]

For once I agree with everything RSJ said. I'm with RSJ.

[JeffreyY]

From the individualist perspective, isn't any corporation or partnership a cartel? How do you determine whether corporations or unions are worse for efficiency?

[Rob]

"The distinction between the individual and the collective good is what's at issue, not the distinction between the good of one group of individuals and another group of individuals."

Maybe I'm not reading the post correctly, but are you arguing that distribution of income doesn't matter? Inequality doesn't matter? Without even the "compensate the losers" argument?

I have to agree with the other commenters, I think.

[Min]

Hmmm.

Suppose that we have two players. A zero sum game has a value, but trade is non-zero. So trade is a win-win. OTOH, I do not see how we can determine that both are paid too little or too much.

Now suppose that we add a third player. That introduces the factor of competition. IIUC, the claim now is that, if the players are comparable in ability and resources, the general welfare increases by more than 50%, unless two players form a coalition against the other. Then the gain of the two from the coalition is less than the loss to the odd man out. The increase in general welfare is plausible, but not the overall loss from coalition forming.

Now humans form coalitions, friendships, partnerships, associations, etc. But relative advantage is not the only reason for doing so. They also form communities. Focus on only trade may lead to a conclusion that you do not draw, that competition is more beneficial than cooperation.

The negatives of competition are obvious, reflected in sayings like, "It's a jungle out there," "It's a dog eat dog world," and "It's a rat race." Yet these negatives do not seem to be included in your equations. Nor do I see any evidence of the positive cooperative values of coalition building.

Now let us look at the three player case with inequality, where one person commands more material resources than the others. Can we really say that the general welfare is reduced if the other two form a coalition? If they compete with each other, the third has an even greater advantage than his natural material advantage, but if they cooperate, they can offset that advantage to some extent. Because they have relatively little to start with, their material gain from the coalition is plausibly greater than the loss to the third. (As with the moral value of money.) Furthermore, from the aspect of the three as a community, the one benefits from a certain minimal welfare of the two. (And probably more than he thinks. The temptation of a master-slave relationship can blind a person to the reciprocal value of the general welfare.)

Finally, an example I found amusing in real life: Three persons with inequality where one has a relative material disadvantage, and the other two are roughly equal. The two top dogs are in perpetual competition with each other. The materially weak one not only benefited from their competition, he was able to form temporary coalitions with one or the other, and so wielded real power. ;)

[Min]

Correction:

I wrote: "their material gain from the coalition is plausibly greater than the loss to the third."

I meant, "the value of their material gain from the coalition is plausibly greater than the value of the loss to the third."

[cives]

I wonder how the idea would look like if you make the -plausible- assumption that firms have some monopsony power.

I say plausible because the same microfoundations that apply to justify monpoly power, usually apply on the other side of the market- search frictions, heterogeinity, etc...

[Nick Rowe]

Thought this post might raise a few hackles!

I'm waiting for the New Keynesian macroeconomists to chime in. They will agree with everything, except the title (which is obviously false), and the conclusion.

For example: RSJ: "What bothers me is that you are taking a highly stylized model that ignores dynamics, investment, heterogeneity, finance, the endogeneity of productivity, and concluding that you "know" that "people" are overpaid."

And it's exactly the same stylised model, which *assumes* that each individual is overpaid, that New Keynesian macroeconomists use to explain why monetary policy has real effects, and why booms are good and recessions are bad. And they don't complain about these same assumptions when they are used to draw conclusions they like. But all of a sudden, when you get a conclusion you don't like, these assumptions are problematic.

You missed the point of the post. Or don't want to see it.

cives: "I wonder how the idea would look like if you make the -plausible- assumption that firms have some monopsony power."

I built a model like that once, and actually published it somewhere. I realised afterwards that it was wildly implausible. Because the model predicts there will be zero unemployment. All firms will be trying to hire more workers all the time. Any unemployed worker would immediately receive multiple job offers. Monopolists always want to sell more at the current price (wage). Monopsonists always want to buy more at the current price (wage). Which of the two assumptions looks more plausible? Monopoly, clearly.

[Nick Rowe]

jeffreyY: "From the individualist perspective, isn't any corporation or partnership a cartel? How do you determine whether corporations or unions are worse for efficiency?"

Some corporations and partnerships are cartels. They exist in order to increase market power. How do I determine whether unions or other cartels are worse for efficiency? I don't. They are both bad for efficiency.

[anon]

Very interesting. The basic idea (one you have discussed here before, I think) is that when a monopolistically competitive supplier lowers her relative prices, the consumer faces a relaxed income constraint and can buy more from other suppliers, thus increasing their profits. But AIUI the aggregate demand curve is not perfectly horizontal, because some of the increased demand will "leak" into input costs, rather than markups. (e.g., perfectly competitve goods, etc.)

What is the monetary authorites' best tradeoff here? Clearly you can't have a "surprise" monetary expansion all the time, but what if you expect to have one with low probability, and have a slightly tighter policy at all other times? Do the negatives outweigh the positives?

[Nick Rowe]

anon; thanks.

here are two thought-experiments I find helpful (maybe one or both will work for you):

1. Cournot. Nash equilibrium with each person setting a quantity, taking others' quantities as given. Starting in Nash equilibrium, if one person increases quantity the loss in that person's utlility will be second order of smalls (since she is starting at the utility-maximising point and the objective function is smooth). But everyone else will consume more of her output and so will have a first order of smalls gain in utility.

2. Bertrand. Nash equilibrium with each person setting a money price, taking each other person's price as given. We have to think of a monetary exchange economy here, because there are n players and so n prices, but only n-1 relative prices. And so we have to bring in the monetary policy reaction function. Assume it's a price-level targeting central bank. Starting in Nash equilibrium, if one person lowers her money price (second order of smalls loss for her), the price level falls, and so the central bank will loosen monetary policy to increase AD, so real output expands, and all gain.

There is no hope for a policy of surprise monetary policy, under RE. The usual implication is to avoid surprises to AD, by offsetting other shocks.

[Nick Rowe]

Min: in a standard NK model, a coalition *of the whole* will make everyone better off. It's very Hobbesian like that. Each agrees to give up his own monopoly power, in exchange for others doing the same. "One Big Union" will try to force all the individual unions to lower wages/prices. "Corporatism" is how it's usually described (notwithstanding the fascist echoes). IIRC, something like this happened in Ireland (and elsewhere) in the 1980's? Can't remember the details. But the main ingredients are a "social compact" where the central union tries to prevent wages rising, and the central bank expands monetary policy to increase output and employment with no need to worry about inflation.

[anon]

"Min: in a standard NK model, a coalition *of the whole* will make everyone better off. It's very Hobbesian like that ... 'Corporatism' is how it's usually described (notwithstanding the fascist echoes). IIRC, something like this happened in Ireland (and elsewhere) in the 1980's?"

It happens in Western Europe and the PRC as well. Depending on who you ask, it's referred to as either "market socialist economy", or "socialism with Chinese characteristics".

[Sandwichman]

Devil's advocate, eh?

[John FitzGerald]

Cartels ain't going away. Governments seem devoted to promoting bigger ones, and the press to ignoring them.

[miep]

Corporatist wage restraint is common in western Europe. The srongest example at the moment are the Netherlands. Ireland social pacts are a rather weak one. Half Europe had such pacts, most had limited impact. In the past, Austria was the most corporatist one. They had some pushback since the neolib (FPÖ) gained more and more power up to the point where they were in government for 8 of the last 10 years. Corporatism worked very good against unemployment in the 70ths during the oil crisis.

The new competitive corporatism had less stellar result. Very recently it looks like countries with more centraliced wage bargaining came much better through the financial crisis. This is important political context.

[Min]

Nick Rowe: "in a standard NK model, a coalition *of the whole* will make everyone better off."

Yes, that seemed to follow.

Nick: "Each agrees to give up his own monopoly power, in exchange for others doing the same. "One Big Union" will try to force all the individual unions to lower wages/prices."

Or One Big Gov't?

The point of lower wages/prices is to spur output, right? More goods and services for you and me?

Nick: " "Corporatism" is how it's usually described (notwithstanding the fascist echoes). IIRC, something like this happened in Ireland (and elsewhere) in the 1980's?"

I am reminded of Japan when I was young. The Japanese web of responsibility took care of output. People wanted to be good worker bees. (My American viewpoint is showing. ;)) Corporations were quite loyal to their workers, even providing sinecures. No need for union featherbedding. Underperforming workers were not regarded as shirkers, as indeed, they were not. (With some exceptions, OC.) Even if you could not afford to join a country club, you could use the company driving range on the roof or, on occasion, play on the company membership. As an American, the whole thing struck me as rather feudal.

Thanks, Nick. :)

[Min]

Another thing:

Nick Rowe: "in a standard NK model, a coalition *of the whole* will make everyone better off."

So then from that perspective the problem with the minimum wage is that it does not cover everybody? E. g., it will cause some unemployment. An answer to that might be Bill Mitchell's Job Guarantee? You don't need a minimum wage because it is implicit in the guaranteed jobs.

[Daniel I. Harris]

The model makes perfect sense to me, and I like it. I have two concerns, though:

1. Unions make everybody worse off in that we can consume fewer goods. But the tradeoff is more leisure, as our work is less demanded, is it not? Personally, I'll take the leisure--more time to learn economics!

2. What about situations like the 19th century, when the owners of capital had a disproportionate share of income? I agree that unions reduced output. I also agree that unions reduced investment in factories as they reduced the ROI. I would even accept that the huge ROI on factories would have eventually forced a rise in workers' salaries as factory workers became scarcer. I also accept that many of these problems were the result of really bad protectionist policy (i.e. The corn laws). I would also accept that a better solution would have been a GAI. But living conditions were atrocious, the government was unsympathetic, and unions forced a dramatic improvement in the lives of the poor. Long term increases were coming but far off, and the conditions needed ameliorating immediately. A reduction in overall consumption and efficiency was worth it.

[RSJ]

"And it's exactly the same stylised model, which *assumes* that each individual is overpaid, that New Keynesian macroeconomists use to explain why monetary policy has real effects, and why booms are good and recessions are bad. And they don't complain about these same assumptions when they are used to draw conclusions they like. But all of a sudden, when you get a conclusion you don't like, these assumptions are problematic."

Hmm, no, I disagree about the potency of monetary policy, too, and have complained about it to the point of ridicule, on several occasions. I'm not trying to debate all the money multiplier stuff now, but you did mention corporatism, and a good example of that is the Wassenaar agreement, which was a (successful) attempt at disinflation without a central bank induced credit crunch.

Anyways, I understand the point -- it's a very simple point, and a nice argument. I just don't think it has much to do with booms or busts in our economy, nor with whether or not people are overpaid or underpaid; one reason for that is the BLS study I cited -- I'm really curious as to how you reconcile that study with your model.

If you would have just made the coordination failure argument as a way to think about monopolistic competition or to illustrate certain types of games, then it would be a great educational post. And a different economist would tell the exact same story, with the same educational value, and then point out how it contradicts the wage-productivity relationship we see, and list that as a limitation of the model. We have data about how much people are paid and how much they produce. Even though this data is imperfect, it doesn't correspond well to the story you tell.

It's possible to see that and at the same time understand the story.

[paine]

professor rowe:

its a very good thing i come here so rarely
my affection for abstraction always suffers
btw
are you ever both relevant and right
err ....while still "doing economics " ??

this mind gaming
never exceeds
the boys with toys level
and of the most pocket protector variety

"since 18 "

indeed

duels with plastic pistols

[miep]

Japan is considered a corporatist country (Schmitter definiton, not Ron Paul definition) by some, albeit an unusual special case - corporatism without labor.

[Lord]

So there is nothing wrong with your model other than it does not accord well with the world. Firms exist for presumably some efficiency reason that may also apply to unions, etc., like Krugman's post on the glories of central planning.

[Sandwichman]

Nick,

When you say "paid too much" and "paid too little" are you referring to wage rates, aggregate income, per capita aggregate income in proportion of effort expended or to some vague floating notion of getting paid that can refer to one thing at one time and another in some other circumstance? If you start out from a vague enough notion of getting paid, all the subsequent cosmetics and costume changes don't contribute any clarity or precision, only deeper obfuscation.

There are not one but at least two plausible interpretations of your "coalition of the whole." One of them would follow John Maurice Clark's analysis of social overhead costs toward a holistic, social-accounting model. The other subscribes to a dualistic "consumer sovereignty" principle, as enunciated by William H. Hutt. The former integrates consumers and producers into a unified complex whole, the latter explicitly subordinates "the producer" to "the consumer," without distinguishing between labour and capital in the role of producer -- presumably they are both equally subservient to the will of the consumer.

The consumer sovereignty position is grounded in very basic accounting errors, some of which appear suspiciously like deliberate evasions.

[Luis H Arroyo]

very good, indeed.
Nick, is not there a problem of imperfect information? that is, is not the prisioner´s dilemma a problem of lack of information? I mean, is not always a problem of incompleted information in all markets?
Is not that the nuclear reason for:
"If all prices are sticky, then an expansionary monetary policy, which increases aggregate demand, increases output and makes everyone better off. That's why booms are good, because it brings the economy closer to the competitive equilibrium. And why recessions are bad, because they take the economy further away from the competitive equilibrium."
If everybody would know the reaction of the neighbors, everyone would search the equilibrum wage.
But never everyone will know that: it´s literally imposible (a joke that Hayek was one of the first in talk about the impossibility of complete information in each and everyone).
I mean: is lack of information is = All against All?

[Luis H Arroyo]

Sorry:

(a joke that Hayek was one of the first in talkING about the impossibility of complete information in each and everyone)

[Nick Rowe]

Sandwichman: "Devil's advocate, eh?" I'm not sure of exactly the right words, but I don't think I'm playing devil's advocate here. "Irony" seems to be closer to the mark. The post title, being obviously false, is clearly ironical. What I'm doing is pointing out a prime facie internal inconsistency in the belief system of many New Keynesians, who are nevertheless pro-union. I'm wondering how they reconcile their pro-union stance with their New Keynesianism. (Not that it's impossible, for any theorist with enough ingenuity; but there is a definite theoretical tension there.) I'm thinking especially of some of the cousins, and their views on Wisconsin. European "corporatists" seem more consistently New Keynesian.

miep: thanks. That's exactly the sort of examples I had in mind, but couldn't remember clearly enough. The logic of the NK model drives you either to a corporatist solution, with centralised wage (and price?) setting, or to trying to weaken the power of individual unions. One extreme or the other. And the centralised wage-setter in a NK model would try to keep wages *down* -- below the level that individual unions would seek in Nash equilibrium -- because one union's wage increase is another union's cost increase.

Min: "So then from that perspective the problem with the minimum wage is that it does not cover everybody? E. g., it will cause some unemployment."

No, the problem is that it is a *minimum* wage. What is needed (according to the model) is a *maximum* wage (or price), that covers everyone. You need a maximum wage to reduce unemployment. And it would have to be individually-tailored (according to demand and supply for each different type of worker) so that it was binding everywhere. The government would lower the maximum wage for any type of labour that was in excess supply (that had positive unemployment). That's the logic of the model.

(Bill Mitchell's proposal is like a gold standard system, except labour replaces gold. And the (only?) problem with Bill's proposal is that gold is homogenous, while labour is heterogenous. So in equilibrium only the very worst labour would be employed in Bill's projects. Which is a pity, because Bill's proposal is very neat theoretically. A closer analogy would be a gold standard, in a country where unemployed workers could pan for gold.)

[Nick Rowe]

Danial: Thanks. "1. Unions make everybody worse off in that we can consume fewer goods. But the tradeoff is more leisure, as our work is less demanded, is it not? Personally, I'll take the leisure--more time to learn economics!"

The logic of the model is that demand prices will be above supply prices everywhere, because of monopolistic power (market power of sellers). Which means the value of "leisure", at the margin, is less than the value of the goods it could produce. And remember, the "leisure" we are talking about is typically the "leisure" of unemployed workers. The extra leisure is not shared around equally; it's "enjoyed" by the outsiders -- the involuntarily unemployed, who would much rather have less leisure and more consumption.

2. You could be a NK and still defend unions by arguing that unions improve the distribution of income, despite the aggregate loss. But, to be consistent, you could only defend unions for those with below average incomes. (My guess is that the Wisconsin teachers everybody down south is so riled about have above average incomes?) And yes, some sort of GAI would presumably be both more fair and more efficient.

RSJ: "Hmm, no, I disagree about the potency of monetary policy, too,..." this point is really orthogonal to the question of the relative efficacy of monetary and fiscal policy. Most NK models have only monetary policy. But I could replace my words "monetary policy" with "monetary and/or fiscal policy", or simply "AD policy", and what I said about NK models would still be true.

I may have a look at that wage productivity link later. Maybe it does or does not count as evidence against the NK model. But this post is really about whether those who believe in the NK model of booms and busts, and at the same time are pro-union, can reconcile their beliefs. (I'm teasing them just a little, and so far none have bitten, but maybe they will. Just hope they don't bite me too hard!)

paine:
Sometimes,
just sometimes,
you have a point
that is
both understandable
and good.
But not
today.
Pity.

[Nick Rowe]

Lord: maybe unions, like firms, create some efficiency benefits. But Paul Krugman also believes, like me, that the world we live in is one of generalised excess supply, for most goods (including labour) most of the time. The demand side of the market is the short side of the market. Sellers have market power in most markets. If you believe that, as we both do, and you accept the NK view of booms being good and recessions being bad, as we both do, then it's not easy to reconclie that view with one that wants to strengthen the market power of labour unions.

Sandwichman: "When you say "paid too much" and "paid too little" are you referring to wage rates, aggregate income,...."

Good point. I wasn't sufficiently clear.

At the individual level, each individual sets price (wage) too high, and has income too high. By setting his relative price above the competitive equilibrium, he increases his own real income, at others' expense.

At the aggregate level, relative prices are not too high (mathematically they can't be, since they must be one on average), but real incomes are too low.

You lost me on your second point. In a macromodel, everyone is both producer and consumer.

Luis: in an NK model in full equilibrium, the underlying problem is not imperfect information per se, but imperfect competition. (Though the imperfect competition might in part be magnified by imperfect information, if buyers are unaware of competing suppliers, but this is not in the NK model). Sticky prices don't really come from imperfect information either, in NK models. It's just assumed that it's costly to change prices. (Though, again, there may be some underlying imperfect information story of those costs, but it's not in the model).

[Simon van Norden]

Nick, is there any particular reason you single out unions for mention as an example of how reducing price elasticity is A Bad Thing? I'm frequently struck these days by how much effort is expended (via loyalty programs, "bundling" discounts, non-transferability of contracts, niche-marketing, efforts to build "brand loyalty", etc.) to make the demand for goods and services less elastic at the level of the individual firm.

Do you think union power is a bigger macroeconomic problem? Or am I reading too much into a simple example?

[Nick Rowe]

Simon: you are maybe reading too much into it. I'm against all those frequent flyer scams too. There is nothing *theoretically* special about unions in reducing price elasticity. Maybe economically more important in *some* countries, since labour, and unemployed labour, is so economically important.

My main reason for singling out unions is that that's where some people's views become inconsistent. I can't think of any New Keynesian macroeconomists who strongly support other "loyalty programs" like trading stamps and frequent flyer points.

Memo to self: someday I must finish writing this post. Unions etc. are really just epiphenomena. At the root is a distinction between buyers and sellers. Sellers have market power; buyers don't. But that very distinction makes no sense in a barter economy. In barter, when you sell something you buy something at the same time. Bertrand competition only makes sense in a monetary exchange economy, because there are only n-1 relative prices to set. Because money is homogenous, each seller of money is in perfect competition with every other seller of money.

[miep]

Dont think there is any meaningfull example for centraliced price seting anymore. Switzerland has some weird milk cartell that might be considered a sector corporatist system. Austria abolished centalicediced price setting long ago.
" One extreme or the other." Its called the hump shape hypothesis. Tálos suggests this meta paper refutes the thesis (p27):
http://siteresources.worldbank.org/SOCIALPROTECTION/Resources/SP-Discussion-papers/Labor-Market-DP/0120.pdf

[Sandwichman]

"You lost me on your second point. In a macromodel, everyone is both producer and consumer."

My second point is that there is more than one way to be "both producer and consumer". Are the two roles integrated or is one of them subordinated to the other? If the model in question equates the "leisure" of the unemployed person to the marginal value of leisure of the employed, then I would describe that model as non-integrative (if not outright absurd).

It's possible to have both high-wage and low-wage versions of the non-integrative model. The two versions foster the semblance of a "debate," which nevertheless excludes the social overhead cost perspective of Clark (and Veblen).

[thomas]

Nick: " "Corporatism" is how it's usually described (notwithstanding the fascist echoes). IIRC, something like this happened in Ireland (and elsewhere) in the 1980's?"

Australia's Prices and Incomes Accord is another example. According to John Quiggin in Zombie Economics, the Netherlands also had an agreement of this sort at the same time.

[OGT]

I am trying to figure out why you're description seems so different than Krugman's. Perhaps, because he's (partially)abandoned the Rep Agent model?

http://krugman.blogs.nytimes.com/2011/01/16/wages-and-employment-again-wonkish/

Also, what if some people are perfectly competitive price takers, say migrant grape pickers, Wal-mart employees, and so on, while some others like Econ professors are monopolistically competitive? Which seems more realistic (not that that has much to do with modeling). Would unionization have some social benefits then?

[liberal_usa]

"I built a model like that once, and actually published it somewhere. I realised afterwards that it was wildly implausible. Because the model predicts there will be zero unemployment. All firms will be trying to hire more workers all the time. Any unemployed worker would immediately receive multiple job offers. Monopolists always want to sell more at the current price (wage). Monopsonists always want to buy more at the current price (wage)."

Nick- I found this paper, though Sciencedirect is down for maintenance, so I can't read it yet.

I don't get intuitively why this scenario would cause zero unemployment. Employers have monopsony power in labor markets over workers. The employers offer fewer vacancies than the equilibrium level as this reduces the wage. Yes, employers would prefer to sell more at this level, just like any monopolist would. But they can't, they would need to raise their wages to hire more. Workers would not get multiple offers, the number of offers would be much smaller than the equilibrium level of offers in a competitive market.

I think your model assumes a totally elastic labor supply curve at the monopsonist wage equilibrium. I dont' see how else you get to zero unemployment with monopsonies. Obviously there can't be zero equilibrium in a monopsony- the equilibrium labor supply is lower than a competitive market, so if the competitive market has non-negative unemployment, a monopsonistic one must as well. Countervailing the bargaining power of management would move the market closer to equilibrium and thus increase total output.

[Andy Harless]

It depends on how you define a New Keynesian, I suppose. The aggregate demand externality due to monopolistic competition may be the critical feature of New Keynesian models, but most New Keynesians are interested in models that have other important features as well, and some of those features may look better for unions. For example, Delong and Summers argued that wage and price flexibility could be destabilizing due the effect of anticipated deflation on the real interest rate. And then there is the recent Eggertsson-Krugman paper which makes a similar argument via the debt-deflation mechanism.

[Determinant]

RSJ's BLS article highlights an interesting point: "dual realities", aka the difference between the GDP Deflator and the CPI Index. I think this throws an interesting wrench into your argument, Nick.

Any thoughts?

[anon]

Andy Harless, shouldn't expected deflation be seen as a monetary problem and not a real one?

Perhaps wage/price flexibility makes incomes more volatile and defaults more likely, so the second paper still applies. But the straightforward solution is to index debt contracts on overall economic conditions, so that relief is available in bad times. AFAICT, there is no externality here, so this is at least as effective as the obvious government solution (mandated debt forgiveness).

[verbatim]

Following up on RSJ - different models are good for different purposes. A macro model written to study the impact of financial intermediation is probably not suitable for understanding the impact of migration. Every model needs to make assumptions and in order to be parsimonious the typical approach is to assume away complexities on dimensions that are not directly related to the question being addressed. It appears to me that standard New Keynesian models are good for studying monetary policy but not so good for studying issues related to inequality, bargaining power, endogenous productivity, etc.

I believe that the preference ordering of economics models depends critically on the specifics of the question being addressed. To apply the same model or, more specifically, to make the same assumptions in order to answer every question is obsessive/close-minded.

I have learned as much from your blog posts (thank you!) as I have from reading textbooks and journal articles. I also agree that unions maybe sub-optimal. However, your attempt to use the NK model to answer this question seems far-fetched.

On a separate note could you in a future post discuss the role of the fundamental theorems of welfare in new keynesian (or even RBC and new monetarist) models? How critical is the complete markets assumption? How does it relate to the representative agent framework? How is welfare defined in heterogeneous-agent models (are we still alluding to the pareto-optimality of CE)? In typical models are we just adding the utilities of different groups? What does it really mean to say that the planners problem with transfers can be implemented as CE if lump-sum taxes and transfers are politically impossible (or endogenous). I have seen various descriptions, but I would really value your lucid and intuitive explanation (and I can learn not only from your post but also from the comments it generates). And thank you again for all that you've taught me.

[verbatim]

I meant to say following up on Andy Harless (not RSJ).

[david]

My impression was that NK models tended to presume that labor markets clear, even in the short run, and despite the presence of labor wage contracts - e.g., here is the New Keynesian Greg Mankiw describing this presumption as an advantage:

[The] cyclical behavior of the real wage does not appear consistent with the model incorporating a predetermined nominal wage and movements along a standard downward-sloping labor demand schedule. [...] These .. problems with the view emphasizing the stickiness of nominal wages has turned the attention of Keynesian macroeconomists in the 1980s away from the labor market and toward the goods market. A "new Keynesian" view has been emerging ... According to this view, the problem in a recession is not that labor costs are too high but that sales are too low.

[...]

[The] New Keynesian view does not imply a countercyclical nominal wage. Once price rigidity is introduced as an important element to explain the response of the economy to changes in aggregate demand, real wages can be procyclical or acyclical. Moreover, if price rigidity is combined with the view that observed wages are merely installment payments, one can obtain Keynesian results while leaving the path of wages completely indeterminate and completely irrelevant.

Very unambiguous. This was in 1988, of course, and it may be that NK modelers have stopped presuming competitive labor. But my impression is that invocations of the canonical NK model still do generally incorporate flexible labor markets.

If you wish to be cynical, you can read a political quid pro quo into it - the economist right stopped screaming about excessively high wages, and the economist left stopped screaming about redistributive effects of macro policies ;)

[david]

Okay, so reading the comments again, it is clear that your impression of NK models tend to have sticky wages as a major element. Do you have citations for this? Isn't the canonical three-equation NK model just have staggered Calvo price-setting with expectations (to generate the expectations-augmented Phillips curve), an IS curve, and a central-bank reaction function? There's no wage-setting in there. Where did that impression come from?

There are, of course, prominent New Keynesians who don't adopt pro-union views (Summers, Mankiw, etc.). And there are New Keynesians who do adopt pro-union views. This is completely unsurprising, since New Keynesian economics does not hold short-run wage stickiness or wage distribution as key canonical elements, which are exactly the issues involve unions, so the NK umbrella can cover pro- and anti-union thinkers.

There's no need to believe that unions bargain the aggregate wage level above efficient levels to support or oppose unions, of course; on the pro- side one could claim that they redistribute wages from high earners to low earners, or that they increase long-run efficiency by reducing turnover/labor monopsony/etc. These sorts of arguments seem common among the Yglesias-type crowd. I trust you already know the anti- arguments.

[RSJ]

Nick,

"I built a model like that once, and actually published it somewhere. I realised afterwards that it was wildly implausible. Because the model predicts there will be zero unemployment."

You don't need to think in terms of monopsony. Productivity is only (objectively) defined for the the industry or firm, not for the occupation -- which is why the BLS doesn't publish occupation productivity data:

"There are also conceptual obstacles to disaggregating these national measures. For example, the output of a factory may require both white-collar and blue-collar inputs, and it is therefore unclear how to allocate the output to the two groups separately."

And in this ambiguity rests the power struggles among the occupations, which are carried out by unions, management and professional organizations. Unions have historically been occupation-based. In an environment with specialization of labor, one occupation is not substitutable for another, and so the production function, in terms of occupations, looks like min(A, B), and so the relative labor share of each occupation is undefined -- it is left to power relationships and bargaining. Now you cannot argue that ideal state will be "no power", as someone needs to decide on the distribution of payouts for each occupational input.

A decrease in the wage share of one occupation matched by an increase in the wage of another wont cause an increase in hiring.

The firm will not be able to hire more workers because that would drive the MPK below the market rate.

As an example, think of universities hiring fewer tenured track professors and a larger number of adjuncts. The adjuncts are paid much less, but tuition does not fall. If we assume output does not fall either, then what is happening is just an intra-firm shift.

If the adjuncts were to form a union and gain in power, then they could take back some of the salary gains experienced by the top professors while keeping overall university costs and output the same.

You don't see these effects in a simple exchange model of production that doesn't have occupations or that assumes that each occupation has a well defined marginal product.

I'm not even aware of a macro model in which "management" is listed as a factor of production. Of course those models don't get unions right, as they can't even model what unions do, let alone determine whether what they do is helpful or hurtful.

[Min]

BTW, this theme is prominent in this blog entry at Crooked Timber, in the context of professional sports: http://crookedtimber.org/2011/02/27/fair-play/#more-19101

[miep]

Strong occupation based unions are not the norm across the world. Industry based unions usually do better for society as a whole, less walkouts, lower wage differentials, less danger to be destroyed altogether.

[Nick Rowe]

OGT: "I am trying to figure out why you're description seems so different than Krugman's. Perhaps, because he's (partially)abandoned the Rep Agent model?"

Because in that post, PK is assuming the economy is stuck in a liquidity trap, and that monetary policy can't work to increase AD. He's talking about a very special case of the same model. I'm talking about the general case, where monetary and/or fiscal policy can be used to shift the AD curve, and so the constraint is inflation.

Andy: same response. If there's a problem with shifting the AD curve, then we are in a special case of the model, where the effects of falling wages and prices may be different, because they can't be offset by monetary policy.

liberal-usa and david:

Draw a standard supply and demand curve diagram. Consider 2 cases:

1. The price is above the competitive equilibrium. Sellers have market power. Monopoly. Quantity actually sold is demand-determined. There is excess supply. Buyers are buying as much as they want at that price. Sellers would always like to sell more at the going price, but can't find a willing buyer. If the price is sticky, then fluctuations in demand will cause equivalent fluctuations in quantity sold. Fluctuations in supply have zero effect on quantity sold, and only change the amount of excess supply.

2. The price is below the competitive equilibrium. Buyers have market power. Monopsony. Quantity actually sold is supply-determined. There is excess demand. Sellers are selling as much as they want at that price. Buyers would always like to buy more at that price, but can't find a willing seller. If the price is sticky, then fluctuations in supply will cause equivalent fluctuations in quantity sold. Fluctuations in demand have zero effect on quantity sold, and only change the amount of excess demand.

Ask yourself whether 1 or 2 best describes the world we live in? The key insight of NK macro is that 1 best describes the world we live in. It best describes the market for widgets; it best describes the market for labour.

Yes, you can always build a fake NK model in which the market for widgets is like 1, but the market for labour is like 2. (As I did once). But doing so is both wildly implausible, and totally violates the whole insight of NK macro. Since labour is, 70% (whatever) of GDP, you can't build a NK model where 70% of the model violates 1 above, and still call it NK. The aggregate labour demand curve is approximately horizontal. Non-human capital, and land, and the resultant diminishing marginal product of labour, are small beer.

In fact, my preferred NK model is a yeoman farmer economy. There are people, selling the fruits of their labour and land and capital (human and non-human), and money. Firms are a minor complication.

[Nick Rowe]

RSJ: you are still thinking in terms of the Us vs Them struggle for revenue shares *within* each factory. The factory as a whole faces a downward-sloping demand curve for its output. Now think about the struggle of All factories against All other factories. That's the NK vision.

If the individual factory has a horizontal LR MC curve, and a demand curve with constant elasticity, it can and will pass on any cost increase with an equivalent percentage increase in its relative output price. But factories as a whole cannot do this, because the aggregate relative price is always one.

[White Rabbit]

Nick:

In your previous post shouldn't "buyers have market power" and "sellers have market power" be switched?

When there's excess supply it's buyer's intent that controls trade. I.e. buyers have market power (and set prices).

When there's excess demand it' seller's intent that controls trade. I.e. sellers have market power (and set prices).

Also, there can obviously be a mix of the two: for physical resources and monopolized products/services I'd say it's sellers that control right now, for all other complex products and services it's buyers.

Or have I misunderstood your point?

[White Rabbit]

Nick:

Ignore my comment - I got it wrong.

[Shan]

I'm no economist, but I can't quite bring myself to accept the idea that the demand curve for labour could ever be perfectly horizontal like that in the real world. Labour productivity isn't some exogenous constant. In your "yeoman farmer economy", you can't grow more food in the aggregate by just hiring more workers and increasing other expenses proportionally. You'd be expanding into progressively more marginal farm land, so rising output means lower productivity and the demand curve slopes downwards. Reminds me of the well-known saying in the computer software business, which notes that "assigning more programmers to a project running behind schedule will make it even later."

[Nick Rowe]

White Rabbit: Phew!

Shan: yes, in an agricultural economy, where land is an important factor of production, and in inelastic supply, you would be right. (My yeoman farmer metaphor perhaps wasn't the best). But if land is less important, then there's only labour and capital. And the long run supply of capital is roughly horizontal. If you double employment, savings, investment, and capital will eventually double too, so there's no diminishing returns to labour, in aggregate, in the long run. Roughly speaking. Land is the only thing that causes diminishing returns to labour.

[Nick Rowe]

Shan: you may not be an economist, but somewhere along the line you seem to have come across David Ricardo's model!

[Shan]

Actual farmland is far from the only important factor of production with relatively inelastic supply. Crude oil comes to mind, and that alone seems big enough to have substantial effects on the economy as a whole.

[Shan]

... and thanks for the quick reply, good to know I'm at least not too obviously and completely wrong.

[david]

Nick,

#1 - the canonical New Keynesian model "totally violates the whole insight of NK macro"? Really?

#2 - okay, accepting the abuse in terminology for now... we'll take the core insight of NK macro to be aggregate demand externalities and price-setting power, as you suggest. The basic problem in applying the price-setting argument to labor is with all the empirical issues which the (old) Keynesians faced when they talked up sticky nominal wages: the real wage is not countercylical! The real wage is not countercylical! Scream it from the rooftops! It is, if anything, somewhere between strongly procyclical to acyclical, depending on how you read the data.

So your yeoman farmer economy is attempting to aggregate prices which don't, in reality, move together (widgets and labor). The flaw is the same one which non-Keynesian economists like to fling at Keynesians, namely that the aggregation is hiding the dynamic of interest.

#3 - I'm concerned that you say that your past dalliances with NK modeling incorporate a monopsonistic labor market - for the record, the canonical NK labor market is neoclassical, clearing in the short and long run, not monopsonistic.

[Nick Rowe]

david: "#3 - I'm concerned that you say that your past dalliances with NK modeling incorporate a monopsonistic labor market - for the record, the canonical NK labor market is neoclassical, clearing in the short and long run, not monopsonistic."

And the prediction of that "NK" model is: zero excess supply of labour (no involuntary unemployment; very strongly pro-cyclical real wages (the economy is moving along a very inelastic labour supply curve); recessions only cause employment to fall because real wages fall so workers choose to quit work and sit at home??

"The basic problem in applying the price-setting argument to labor is with all the empirical issues which the (old) Keynesians faced when they talked up sticky nominal wages: the real wage is not countercylical!"

Nope. If nominal wages and nominal prices are equally sticky (seems a roughly plausible assumption), then you get acyclical real wages out of the model when there is monopoly power in both output and labour markets. Plus you get an excess supply of labour (involuntary unemployment) which increases in recessions and decreases in booms. Sounds right to me. If prices are slightly more sticky than nominal wages, then you get slightly procyclical real wages.

[Nick Rowe]

Yep. If the "canonical" NK model has a competitive labour market, and perfectly flexible wages, then the builders of that model wimped out, and threw away the NK baby.

(Of course, if you aren't interested in the labour market per se, and are not interested in the determinants of the natural rates of output income, real wages and employment, and are only interested in business cycle fluctuations in output, then you can ignore the labour market completely. It makes little difference what you assume about the degree of competition in the labour market. Which is presumably how (say) the Bank of Canada's modellers would justify their choice).

[Andy Harless]

A "very special case" perhaps, but arguably a very important one: Delong and Summers (at least in the version of their paper that I read) argue that labor unions actually did help stabilize output during the period after WWII as compared with earlier experience. And it's interesting that the zero interest rate constaint has become binding in the US again now that private sector unions are pretty much out of the picture.

In what you call the general case (by which you mean, I take it, the complementary special case), I'm not sure we're really talking about a macro problem any more. New Keynesians will take the degree of monopoly in the economy as given and argue that shocks can be offset with monetary policy. Would the economy produce more in equilibrium if there were less monopoly power? Of course, but any microeconomist could tell you that. And many of those microeconomists support unions, presumably because they believe unions have benefits that offset the efficiency loss from their monopoly power. (In particular, for example, there is the argument that unions enable workers to express their preferences and thus add informational efficiency.)

And then of course there is the political role of unions. In the US, unions have been important in supporting the Decmocrats, and I suspect that a lot of Democratic economists regard unions as a sort of necessary evil in this respect. Given that the Republicans have now made Ron Paul head of the House Banking Committee and almost tried to block Peter Diamond's Fed confirmation, this is a point that many New Keynesians can probably appreciate.

[paine]

"I'm wondering how they reconcile their pro-union stance with their New Keynesianism"

fiscal policy

[Nick Rowe]

Andy: IIRC (it was a long time ago) DeLong and Summers are arguing that it is the *stickiness* of wages that helps prevent recessions, rather than the *level* of wages. So I expect we are arguing at cross purposes, because I'm thinking about the level of wages/prices.

I don't see why labour unions would increase the natural rate of interest, though (and thus make liquidity traps less likely). If anything, I would have thought that strong unions would reduce the return to investment (ex-post hold-up problems) and so reduce the natural rate of interest.

"Would the economy produce more in equilibrium if there were less monopoly power? Of course, but any microeconomist could tell you that." Well, I think that's a *general* equilibrium question. In that (debatable) sense macro.

Sure. You might consistently be a NK macroeconomist and still support unions, if you think there are offsetting advantages. I expect if I saw a NK macroeconomist say "well, despite this major problem with unions, there are some offsetting benefits, so.." I would give this more credibility.

paine: sorry, but you lost me there. (And I do mean "sorry" in this case, because I don't think you were trying to be obscure; but I just don't get your point about fiscal policy.)

[liberal_usa]

"Ask yourself whether 1[Monopoly] or 2[Monopsony] best describes the world we live in? The key insight of NK macro is that 1 best describes the world we live in. It best describes the market for widgets; it best describes the market for labour."

I agree that product markets are better characterized as demand constrained and exhibiting monopoly power. But with labor markets, I disagree. At least in the US, businesses that hire minimum wage workers like fast food will "always" be hiring outside of recessions. That may be because of turnover, etc., but it always strongly points in the direction of monopsony. Also, unions tend to push up the wages of the unskilled the most, which again is consistent with monopsony, especially among the low skilled who have less bargaining power. Finally, Card and Kruger's results that challenge the effect of minimum wage increases on unemployment, while not uncontroversial, are consistent with monopsony, and definitely inconsistent with monopoly. So I don't think your argument that labor markets are demand determined convinces me.

"Yes, you can always build a fake NK model in which the market for widgets is like 1, but the market for labour is like 2. (As I did once). But doing so is both wildly implausible, and totally violates the whole insight of NK macro. Since labour is, 70% (whatever) of GDP, you can't build a NK model where 70% of the model violates 1 above, and still call it NK."

OK, but just because it's not a NK model doesn't make it wrong. Let's make models where business has upstream and downstream market power- it's a monopsonist for its inputs and a monopolist for its outputs. We don't have to call it New Keynesian- let's call it Robinsonian after Joan Robinson. I still think your point about the inconsistency of many NK economists stands. But I also think that the NK label is too broad and often has nothing to do with Keynes' GT, which doesn't rely on sticky prices or wages nor on on imperfect competition to drive its results, unlike NK models.

I think your point is a good one that New Keynesians should oppose unions to be consistent with their theories about the macroeconomy. I look at it a little differently though- this should signal to New Keynesian two Keynesians two thing- some flaws with their macroeconomic theories and that their theories don't stem from Keynes in any real sense.

[liberal_usa]

The main problem I have with sticky-price based NK models is that sticky prices have never been shown (for me) to have significant enough effects to cause something as large as a recession. Krugman takes the close correspondance between nominal and real nominal exchange rates to be good evidence for New Keynesian theories- for me it's a serious counterargument. Nominal exchange rate volatility combined with sticky prices causes large real exchange rate volatility, instead of no real exchange rate volatility as one would have with total price flexibility. This shows that prices are sticky (what Krugman wants to show), but when you reflect on the implications, this shows that sticky prices don't matter. If they did, countries with fixed exchange rates would have much more stable output and much less of a business cycle. Just think about how much the relative prices of import goods and domestic goods fluctuates just over the course of a single day! That should cause major output fluctuations in a New Keynesian model. But we don't see that- importers just eat that volatility, and keep their prices largely unchanged. If they wanted to pass on changes in their costs to changes in their prices, they could, but it's clearly suboptimal to do that. This doesn't impose a large cost on importers, otherwise this prices volatility would cause trade to collapse in countries with flexible exchange rates. Do you think this to be the case? I certainly don't. Canada's long experience with floating exchange rates combined with large trade flows and stable output doesn't bear this out either.

[Steve Roth]

One of the most instructive posts and threads I've read for a while. Thanks.

re: "One Big Union"/"coalition of the whole."

I would love to hear from those with deeper knowledge and understanding than I (not a high bar):

Imagine a greatly expanded Earned Income Tax Credit, delivered on weekly paychecks (as with withholding). Anyone who works earns a pretty good living -- enough for some leisure time, some nice birthday gifts for the kids, vacations with the family every year or two, etc.

Is that the kind of "big union" that could fundamentally change the "game," break the prisoner's dilemma -- especially regarding all the smaller unions? (Would we need them?)

How do the supply and demand curves play out in that scenario?

I see that Daniel I. Harris goes to the same place: "I would also accept that a better solution would have been a GAI." And that Nick seconds it: "And yes, some sort of GAI would presumably be both more fair and more efficient."

But I'd love to hear a more technical explanation of how it plays out. Nick: worth a post?

[Steve Roth]

Nick: "What is needed (according to the model) is a *maximum* wage (or price), that covers everyone. You need a maximum wage to reduce unemployment. And it would have to be individually-tailored (according to demand and supply for each different type of worker) so that it was binding everywhere. The government would lower the maximum wage for any type of labour that was in excess supply (that had positive unemployment). That's the logic of the model."

Wouldn't a high marginal tax rate constitute such a maximum wage? (But without the necessary individual tailoring?) As under Eisenhower, for instance (91% marginal), so you couldn't make more than two or three million dollars a year (in today's dollars).

Given the excellent economic growth during those periods of high marginal rates, I'm wondering if you haven't really hit on something here.

[Nick Rowe]

liberal_usa: "But with labor markets, I disagree. At least in the US, businesses that hire minimum wage workers like fast food will "always" be hiring outside of recessions."

Actually, in this particular case, I tend to think you are quite possibly (probably?) right. I don't have much first hand experience of this, but I have heard similar stories in the past that confirm what you say, and it has shaded my opinion on this question in the past. If we observe monopsony anywhere, that does tend to look at least a bit like it.

It's sort of like the exception that proves the rule? Where would we be most likely to see monopsony power? In the US, where unions are weakest. In the US, where minimum wages are low. In the fast food industry, where it's presumably(?) hardest for workers to organise, and where the individual restaurants themselves face very elastic demand curves, so their derived demand for labour is also very elastic. Plus, as you say, there's the Card evidence.

I'm not going to say that labour markets are always and everywhere monopolistic. (Nor would I say that output markets are always and everywhere monopolistic.) There are exceptions. But those exceptions tend to make the newspapers. "Man bites dog!" "Labour shortage of nurses!". We know they happen. But the very fact they make the news says they are not typical.

I disagree with your second comment, and agree with Paul Krugman.

"If they did, countries with fixed exchange rates would have much more stable output and much less of a business cycle."

I don't think that follows. It depends on monetary policy, and on *why* the real exchange rate fluctuates. If the real exchange rate fluctuates because of stupid random shocks to monetary policy, then you would be right. But if monetary policy is good and sensible, then fluctuations in real exchange rates and real interest rates are the shock absorbers that are responding correctly to other shocks in AD to help keep output stable.

For example, if an open economy New Keynesian IS curve is Yt= B(real interest rate) + C(real exchange rate) + shock

Then good monetary policy will see both the real interest rate and real exchange rate responding to that shock, to keep Yt growing smoothly. (The real exchange rate should deal with the permanent component of the shock, and the real interest rate the transitory component of the shock, roughly speaking).

[Nick Rowe]

Steve: thanks!

I lean in favour of the GAI, but I'm not really very knowledgeable about it. Here's a post of Stephen's on the subject: http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/11/bi1.html

I don't think high marginal tax rates give the same results as a maximum wage law. A high marginal tax rate would cause low *after tax* wages, but it's the *before tax* wage being too high that is the problem in this case. That's what causes people to demand too little. I haven't actually built this into a model to check my intuition is right, but I think it would just make things worse.

[Steve Roth]

>I lean in favour of the GAI, but I'm not really very knowledgeable about it. Here's a post of Stephen's on the subject: http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/11/bi1.html

Thanks for this. But discussions of a pure GAI don't really address the merits of the EITC, which requires and rewards hence encourages work (and/or hiring, via incidence), and in general seems to have different dynamics (micro and macro). I'll keep looking...

>A high marginal tax rate would cause low *after tax* wages, but it's the *before tax* wage being too high that is the problem in this case. That's what causes people to demand too little. I haven't actually built this into a model to check my intuition is right, but I think it would just make things worse.

Will ponder this at more length. If you ever add it to your model I'd love to see the results.

[Andy Harless]

I don't think it's really a general equilibrium question (until you get to the role of money). You're just aggregating a lot of partial equilibrium results, and AFAIK they're purely additive. Everyone's wages are too high, but only because you've assumed that everyone has monopoly power, so the usual partial equilibrium results about monopoly happen to apply to everyone. The general equilibrium aspect only becomes important when you bring in sticky prices/wages, but as you say, you're talking about the level of wages and not the stickiness. Rather than ask why New Keynesians like unions, you could ask why any economists like unions.

[liberal_usa]

Nick-

Regarding Monopsony vs. Monopoly in labor market, let's chalk it up to half empty/hall full.

Reagrding the exchange rates, my point wasn't really whether exchange rate volatility was optimal, or regarding its relationship with monetary policy, was not really what I was thinking about, and I think your analysis is largely correct.

My point was basically twofold-

(1)that prices in an open economy are clearly sticky, as exchange rate pass through is far from complete

(2) Fact 1 has an insignificant impact on output dynamics, unlike the predictions of most New Keynesian models

I tend to agree with Keynesian/monetarist views of the business cycle (New Keynesian). But I disagree that sticky prices and wages are the way to do it. There needs to be a good way for money to translate to output (non neutrality) without relying on price inflexibility- but I don't know how to achieve that yet. New new Keynesian?

[anon]

"I don't think it's really a general equilibrium question (until you get to the role of money). You're just aggregating a lot of partial equilibrium results, and AFAIK they're purely additive."

Andy Harless, the general equilibrium analysis of monop. comp. is the one which best describes the externality Rowe is talking about in the OP: one supplier lowers prices and expands output--decreasing her own profits by a second-order amount--and this leads to a first-order increase in profits at other suppliers, via consumers' income constraints. A partial equilibrium analysis doesn't make this clear, because it doesn't take income effects into account.

IIRC this is not quite right, because expanding output at one firm can also raise input costs at all other firms. But in the real world, input costs seem to be fairly elastic, so the first effect should dominate.

[RSJ]

": you are still thinking in terms of the Us vs Them struggle for revenue shares *within* each factory."

Yep. That's what unions are about. You are talking about unions, right? But your model doesn't even have any occupations -- how can it shed light on unions?

I tried to expand the model to include some goings on within the firm, so that we could try to describe what unions do, assuming perfect specialization of labor (i.e. no occupation can be substituted with another, so all occupations are perfect complements).

Even if you don't want to assume that degree of specialization, so that the marginal product of each occupation is undefined, you can still assume that the costs of determining the shapley value are prohibitive, so that the distribution of wages for each occupation will not be set by firms running supercomputers doing case analysis, but by bargaining and custom. You might see occupations with a lot of women in them paid less, for example, or occupations that are held in high regard paid more. But you would try to minimize the distortions by having all occupations have equal bargaining power vis-a-vis each other -- hence the existence of unions, which have little to do with the demand curve for the firm's output, and are primarily focused on the struggle over the distribution of labor's share among different occupations.

I guess an equivalent thought experiment would be the following: suppose a firm had three occupations: management, accounting, sales, and laborers. And suppose these three groups collectively bargained to reach the following agreement:

We agree that 65% of the firm's revenue will be paid out to labor, of which 20% will go to management, 10% will go to accounting, and 30% will go to sales, and 40% will go to laborers. Would such an agreement increase unemployment? Now suppose the power relations change, so next year all the numbers are shuffled around. Would this increase or decrease unemployment? Yet we can imagine that in some sense, both of these agreements can't be equally fair, right? Someone must have been overpaid and someone else must have been underpaid, and yet unemployment was unchanged.

In any case, we can agree to disagree, but next time you decide to make an argument about the effects of unions on unemployment, you might want to actually include Labor and Management in your model, so that you have something to say about unions. Then we could kick the model around and decide whether the policy prescriptions are to be believed.

[Patrick]

Further to what RSJ is talking about:

Maybe I've spent too much time in Dilbert world ... Models always seen to assume that the firm maximizes profits mechanistically. But the firms actions are set by management. Management are people too (I can't believe I just said that). People prone to look after their own interests. My experience in firms large and small (8 of them in 11 years) is that any attempt at profit maximization is strictly constrained by individual managers maximization a whole other set of personal 'utility functions' FIRST, then they worry about profits. Maybe.

Some examples: a manager isn't going to fire herself and give her job to the more able superstar underling. In fact, her best strategy is to actively seek the removal of the superstar and replace them with a someone less able.

Or say management as a whole (i.e. the C-suite) makes a dumb decision that reduces profits. If they admits the mistake it can be corrected and profits will increase, but they risk loosing face and having their egos bruised in front of her peers. Some might even risk being fired. What's the best strategy? CYA (Cover Your a**)! To hell with profits; they want to keep their jobs and status.

Management is a union.

[Nick Rowe]

RSJ: yours is a partial equilibrium thought-experiment. You are looking at the conflict within each factory. What you have left out is the conflict between factory and factory. What you are implicitly assuming is that an increase in one group's monopoly power within the factory is exactly offset by a decrease in other groups' monopoly power within the factory, so that the monopoly power of the factory as a whole is unchanged.

Let me try to give you the intuition:

Have a look at my old post:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/01/macroeconomics-with-monopolistic-competition-in-pictures.html

Start in equilibrium. (You can even assume fixed coefficients at the level of the individual firm, if you like, so that the Marginal Cost is the cost of hiring *all* the extra inputs needed to produce one extra unit of output.)

If an increase in union bargaining power raises wages and causes an upward shift in the individual firm's MC curve, in the top diagram, that individual firm will choose a higher price as a result. But that means the macro MC curve in the second diagram will shift up too. So aggregate income falls.

[ptuomov junk]

All economists should hate cartels, labor unions, and minimum wage laws. But New Keynesian modelers who use cartels, labor unions, and minimum wage laws to produce an equilibrium in which cartels, labor unions, and minimum wage laws screw up the world should especially hate cartels, labor unions, and minimum wage laws.

My casual reading of people (Krugman, DeLong, etc.) who blog about economic policy and rely on the NK models to generate their fiscal and monetary policy prescriptions is that they do hate cartels but strongly support labor unions and minimum wage laws.

The most obvious logical conclusion from this is that these bloggers only want to pick the conclusions and policy prescriptions from the NK models that suit their ideology:
- Run high inflation, in reality to create a wealth tax via the taxation of nominal capital gains, under the smoke screen of monetary stimulus and zero bound
- Increase government spending at all times, in recessions under the smoke screen of stimulus and in booms because we can afford it
- Increase income transfers in reality because of purely ideological reasons but ostensibly because of income transfers to their favorite recipients have "higher multipliers."

Then they ignore their NK macro models when they support higher minimum wages and laws that allow labor unions to run labor cartels.

Then they criticize people opposing unions for being motivated by ideology.

Unintentional comedy rating maxes out.

[Nick Rowe]

ptuomov: I see that subtlety is not your strong suit ;-)

I wouldn't go quite that far. People don't always understand all the implications of their views. Sometimes people have conflicting reasons.

[RSJ]

Nick, I get it.

But mine is not a partial equilibrium analysis at all, it's just an analysis with different assumptions.

You are assuming that an increase in union power must come at the expense of an overall increase in the labor share of output.

Why would that be the case?

Historically factor shares in the corporate sector have been relatively stable even as union power first increased and then decreased.

Same for unemployment rates.

You're not going to find a relationship between union power and unemployment, anymore than you will find a relationship between union power and factor shares.

So I think my assumption is fairly reasonable, particularly given that the point of bargaining is to bargain with management, not capital. The rate of interest does not change as a result of bargaining, but management's compensation changes.

And when unions became weaker, the result of this weakening was not lower unemployment or more capital income, but rather higher levels of management compensation, primarily top management compensation.

It's really a simple point, and we can test this point against data, but for some reason it's not getting through..

Arguing against unions reminds me of those who complain about "government". We're not going to get rid of government, so the best we can do is separation of powers, so one branch doesn't get too powerful.

Trying to suppress unions while pretending that management does not exist, or has no power, is an intellectual dead end.

Now perhaps you have a better model for this competition, but your model, if you are going to talk about unions, needs to include at least two occupations that agree to share revenue somehow. A one occupation model, whether general or partial equilibrium, isn't going to capture the effects of bargaining. You need two sides to bargain.

[Nick Rowe]

RSJ: This is how I think of the traditional union model: there's 2 factors, labour and capital. An increase in union power means an increase in labour's share and a decrease in capital's share. And that traditional model is roughly what you have in mind, I think.

"You are assuming that an increase in union power must come at the expense of an overall increase in the labor share of output."

?? No I'm not. That's not what's driving my model.

Here's my simplest model, massively (over-)simplified. There is only labour. So labour's share is 100% by definition.

"Now perhaps you have a better model for this competition, but your model, if you are going to talk about unions, needs to include at least two occupations that agree to share revenue somehow. A one occupation model, whether general or partial equilibrium, isn't going to capture the effects of bargaining. You need two sides to bargain."

No you don't. Or rather, it's union vs customers, where customers are the other unions. So it's union vs union. A warre of all unions against all unions.

Each union can gain at the expense of other unions by raising its price. But when all do this, all lose.

We are still probably arguing at cross-purposes.

[liberal_usa]

ptuomov-

This contradiction arises because Krugman and De Long (and Nick Rowe) accept a theory of New Keynesianism which has nothing to do with Keynes. They're trying to wed a conservative macroeconomic theory with their center-left political beliefs, and it fails. For me, this is the way to understand Nick's excellent point about New Keynesian contradicitons. See this excellent blog post by Waldmann:

http://rjwaldmann.blogspot.com/2010/10/scott-sumner-on-what-keynes-wrote-about.html

Keynes criticizes Pigou's classical theory of unemployment for being New Keynesian. Classical unemployment results from wage and price inflexibility. Keynes *nowhere* assumes price inflexibility, and accepts wage inflexibility as an empirical reality, but not as a source of unemployment, which arise from deficient demand.

The problem here is the New Keynesian moniker, which relies on inflexibility, which has nothing to do with Keynes. In reality, in modern macro we have the New Keynesians, who should be called the New Classicals or New Monetarists, when reliant on price and wage inflexibility. The New Classicals should be called the Short-Run growth theorists, as that is truly their theory- Solow for the business cycle. And as for New Keynesians? You'll know it when you see it- for now, I don't see anyone in modern macro that really stays true to Keynes. (maybe the post-Keynesian, but they're not really in the mainstream)

[anon]

liberal_usa, this post does not mention price rigidity. It mentions imperfect competition, which is needed in any consistent account of deficient demand. If there is no imperfect competition, then all firms are selling their products at marginal cost, and expanding AD doesn't do any good.

Keynes' account of "deficient demand" relies on capital being "idle" and ready to be called into employment when demand expands, thus increasing profits and wages. This is a description of monopolistic competition, which New Keynesians accept. The New Classicals and New Monetarists seem to be more reliant on PC assumptions.

[liberal_usa]

"liberal_usa, this post does not mention price rigidity"

anon, from the post:
"If all prices are sticky, then an expansionary monetary policy, which increases aggregate demand, increases output and makes everyone better off. That's why booms are good, because it brings the economy closer to the competitive equilibrium. And why recessions are bad, because they take the economy further away from the competitive equilibrium."

"It mentions imperfect competition, which is needed in any consistent account of deficient demand."

I disagree- one can have deficient aggregate demand with perfect competition. Why not?

"If there is no imperfect competition, then all firms are selling their products at marginal cost, and expanding AD doesn't do any good."

Yes, they are selling at marginal cost, but at that price=MC, demand is below the level consistent with full employment. Then expanding AD does do good. Why not?

"Keynes' account of "deficient demand" relies on capital being "idle" and ready to be called into employment when demand expands, thus increasing profits and wages. This is a description of monopolistic competition, which New Keynesians accept. "

This is a product of Robinson, that's why NK should be called New Robinsonian. If you can find where Keynes relies on imperfect competition, I'd love to see it. I haven't come across that yet. Keynes is talking about deficient demand under conditions of perfect competition and price flexibility.

"The New Classicals and New Monetarists seem to be more reliant on PC assumptions."

Look at Cole and Ohanian's papers on the New Deal and the NRA- they rely on monopoly power to drive their recessions, just like New Keynesians. And a New Monetarist like Scott Sumner definitely relies on sticky prices and wages, though I guess he doesn't rely on imperfect competition.

How can imperfect competition drive recessions without sticky prices and wages? Let's say the whole economy is characterized by monopolistic competition. Now there's an unexpected decrease in the money supply. How do you get a recession? Why isn't money neutral and why don't prices and wages fall proportionately with money?

[Rick]

Nick,

Honestly?

These conversations are always baffling to me. Wages and Equity are liabilities on the balance sheets of every company. 65% of equity is paid as dividends, not reinvested in companies in the form of new capital. If the idea is that share holders are somehow superior to laborers, ok fine, but what of that logic if the laborers negotiate shares in the company as a defacto raise rather than a direct increase in the hourly rate???

In a very real sense shareholders and laborers are the same.

Its generous to even entertain such a logical stretch as this scenario in which 2 indistinguishable entities redistribute profits amongst themselves causing a magical decrease in employment... or that it even causes a raise in prices, negatively affecting consumers i.e. themselves and other unions.

Willful blindness... I can't believe you are ever serious.

[Rick]

I guess though for your model to work the shareholders/laborers, one and the same person, would be the sole decision makers regarding the spending of the firms capital and profits; thus being a non negotiating union... paying themselves whatever they please (so long as they stay in business)...

So your model is a defacto barter economy of individuals, or groups that act as individuals, all setting their own prices for products and services at a level that they choose to be arbitrarily high, trading with asymmetric information in an imperfectly competitive marketplace to each others detriment.

Your model, your rules, though not exactly useful...

If that's the case though, why even bring up the term union; it doesn't fit and seems a bit malicious?????

[RSJ]

"there's 2 factors, labour and capital. An increase in union power means an increase in labour's share and a decrease in capital's share. And that traditional model is roughly what you have in mind, I think."

OK, that's not what I meant.

"Labour", in terms of industrial relations, means blue collar employees as opposed to supervisory employees. It is illegal for the latter to be a member of a union. So "labour", in the sense you mean, consists of both "Labour" and "Management", in the industrial relations sense.

But unions only represent Labour, not labor. And unions bargain with Management, not capital. I.e. the bargaining process is within labor, with one occupation bargaining with another occupation. You are arguing that this tug of war may spill over into higher output prices.

I tried to make a case for the necessity of this bargaining by arguing that in an environment with high degrees of specialization and a complex production process, it's not feasible to determine the marginal productivity of an occupation. I cited the BLS comment to the same effect. You can talk about the productivity of a firm, or an industry when you aggregate all occupations in each, but you cannot talk about the productivity of a given occupation within a firm or industry, and therefore you cannot argue that occupational wages are determined by productivity.

You can say that one accountant who is more productive than another accountant should be paid more, but you cannot compare the productivity of an accountant to that of technical writer; there is no productivity-specific explanation of why a technical writer would be paid more than an accountant, or a product manager, or a process engineer, etc.

So there has to be some way, other than productivity, to allocate business output among occupations.

Now if you ban all unions, then you will not ban all occupational coalitions, and the wage determination question still must be addressed. And you still have Management which has power because of how we organize production; you still have lawyers and professional organizations that have monopoly power, etc. It is impossible to get rid of all occupational coalitions, and given that constraint, is it beneficial to just eliminate some occupational coalitions? How could you answer that question if your model contains only "labor" and no occupations at all?

I don't support blanket attempts to undermine unions as the result will not be lower output prices, but greater compensation of management and professional organizations -- basically the other occupational coalitions that are still standing will obtain a larger share of business output at the expense of those occupations whose bargaining power you've reduced. I do support attempts to equalize the power of occupational coalitions. So if unions were stronger than management then this would be harmful, just as it would be harmful if the power imbalance was in the other direction.

You argue that if this happens, then the customer will be forced to overpay for output. But the downward sloping demand curves are due to firms selling differentiated products, not due to unions.

The differentiated products represent market power, which is created as a result of investment. The fact that investment permanently continues means that that the returns to capital are permanently higher than the cost of capital, so that, based on marginal considerations, capital appears to be permanently overpaid and the capital stock is constantly being increased. There is a steady stream of net investment, corresponding to a steady stream of market power available to be created as a result of investment. You can think of it as invention if you want.

I don't think this has anything to do with occupational coalitions, which is not to say that you can't create models in which output is lower because of cartels. I just don't think that those models capture our economy very well.

Alternately, you can think of market power as price we must pay in order to have the creation of new markets -- i.e. invention and productivity growth -- and therefore enjoy increasing incomes (over time), just as you view political or extra-economic occupational wage determination as a price we must pay to enjoy the benefits of specialization and highly organized production.

But you need a model with specialization of occupations or endogenous market creation in order to determine whether the benefits outweigh the costs. Thinking of yeoman farmers in a stationary economy isn't going to shed any light on this. Whatever conclusions you reach from such a simple exchange model aren't going to lead to reliable policy prescriptions.

[Nick Rowe]

Rick: "Nick, Honestly? ...These conversations are always baffling to me...Willful blindness... I can't believe you are ever serious....it doesn't fit and seems a bit malicious?????"

OK. Listen up.

One of the things that pro-union people need to damn well understand is that reasonable and thoughtful and moral people can disagree with them. And they have to respect that disagreement, and not respond with stunned incredulity, accusations of malice....(or baseball bats).

I didn't understand the rest of your comment.

[anon]

liberal_usa:

"'If all prices are sticky, then an expansionary monetary policy, which increases aggregate demand, increases output and makes everyone better off.'"

I'm quite sure that Keynes does much the same thing. He describes his model of deficient demand, but takes fixed wages/prices into account when considering its consequences and policy responses.

"I disagree- one can have deficient aggregate demand with perfect competition. Why not?"

Say's law states that there is no "general glut" in perfect competition. Yes, there might be excess demand for money, but only if the price level doesn't adjust.

"Yes, they are selling at marginal cost, but at that price=MC, demand is below the level consistent with full employment. Then expanding AD does do good. Why not?"

If there are unemployed resources, why doesn't their price/income fall? Why don't firms slash prices so that output rises despite depressed demand? That response would be consistent with marginal cost pricing.

"This is a product of Robinson, that's why NK should be called New Robinsonian."

The term "monopolistic competition" may be due to Robinson. But the model of "idle" capital which I sketched above was clearly understood by J.S.Mill ("Of the Influence of Consumption on Production"), and AFAICT by Keynes as well. And it clearly relies on prices being set above marginal cost, which in modern terms is imperfect competition.

Rick, how many laborers want to be residual claimants of the company? Even union laborers? Most unions negotiate fixed wage contracts, not profit shares. Share holders are not superior to laborers, but running the firm is their responsibility. If someone thinks they could manage the firm better than current shareholders, they can launch a takeover bid.

RSJ, we can in fact evaluate the marginal productivity of different occupations. If firms hiring technical writers are paying them more than firms hiring accountants, this means that a technical writer contributes more to the economy than an accountant, so we'd want people to choose technical writing as their career, not accountancy. Occupational coalitions distort this process: if doctors are highly paid but the AMA is unwilling to let me join their club, I cannot become a doctor even if I have stellar qualifications.

[Nick Rowe]

RSJ:

1. The Neoclassical Theory of Value and Distribution does *not* say that factor payments are (generally) *determined by* marginal productivity. (It says they are equal to, and both are co-determined in general equilibrium).

2. As a corollary, fixed coefficients technology (where marginal products are undefined) does *not* (generally) mean that factor payments are indeterminate in the NTVD.

[Nick Rowe]

anon and liberal-usa: since you are both having a good argument, on-topic, and both making good points, I'm just going to sit back and read!

[Lord]

I tend to agree with RSJ on this. For unions to present a problem you must believe they are effective, and not just against management or in capturing increased efficiency, but in raising their wages and the prices of their output above what they would be otherwise. I doubt that they do. I even doubt their political contributions are effective. I see them as much weaker than even firms. The unemployed don't generally stay union and unions must compete against them anyway, there are always other factories.

[RSJ]

Nick,

The effect of the undefined MPP is to remove the MRP floor from wages. You can still have wages be too high as a result of bargaining, but now, they can also be too low. Before, you were saying that they could not be too low, as firms would always want to hire more workers. That is only true if there is just one type of worker, and if MPP is defined for that worker. As soon as you get two occupations, say, L_1 and L_2, with the constraint L_1 = L_2, then

If a firm finds a solution:

p*f(K, L_1, _2) = rK + w_1L_1 + w_2L_2

Then the following is also a (potential) solution:

p*f(K, L_1, _2) = rK + w_2L_1 + w_1L_2

As far as the firm is concerned, it is willing to pay any combination (w_1, w_2) as long as w_1 + w_2 <= pf/L_1 - r(K/L_1)

That means that the ratio of the wages paid to L_1 and L_2 are going to be completely determined by relative bargaining power in the labor market.

As a result, reducing the power of just one occupation -- e.g. unions -- is going to be harmful, since you don't have any guarantees that the other occupations -- whose power you have not reduced -- wont cannibalize its wages.

This is not a theoretical discussion -- in the U.S., we've seen CEOs and top management cannibalize wages on a massive scale.

Anon -- you are assuming the conclusion. The point is that what one is paid is *not* their "contribution to the economy", because marginal physical product of a single occupation is indeterminate.

[Jim Rootham]

@Nick

As long as this is a nice little argument about models with no connection to reality, you can object to unions all you want. Move it out to the real world, where power relations are the key, you deserve the baseball bat between the eyes.

What is missing from this model? Historically, firms want to pay less than marginal rates, it is one of their sources of rents. The use of violence and state power to do that isn't to be seen in the model.

More to your taste, perhaps, is the absence of technological change. The upward pressure on wages that unions create leads to circumstances where it is more profitable to invest in labour saving technology. Also, union shops will tend to have more stable workforces, and learning effects will also generate productivity increases. See here for a business view of unions.

[anon]

"Historically, firms want to pay less than marginal rates, it is one of their sources of rents. The use of violence and state power to do that isn't to be seen in the model."

Yes. Kevin Carson has pointed out a number of state policies which place unionized workers at a disadvantage, compared to a purely free market. Nevertheless, current policies force management to bargain with a _single_ union: regardless of "right to work" laws, workers _cannot_ negotiate separate deals or form competing organizations within the workplace. This is a severe government interference, which makes organized labor much more collusive than it would otherwise be.

[Nick Rowe]

Jim Rootham: "@Nick As long as this is a nice little argument about models with no connection to reality, you can object to unions all you want. Move it out to the real world, where power relations are the key, you deserve the baseball bat between the eyes."

That is exactly the sort of thuggish language and behaviour that is not at all acceptable, either on this blog, or in the real world.

Please retract that immediately, or stop commenting on this blog.

[Mandos]

One of the things that pro-union people need to damn well understand is that reasonable and thoughtful and moral people can disagree with them. And they have to respect that disagreement, and not respond with stunned incredulity, accusations of malice....(or baseball bats).

Unfortunately, the elephant in the room is that there is a relationship of power that underlies this entire discussion: the very wealthy versus the workers. ie, class. For that reason, people who make even well-intentioned arguments against unions must understand why they are viewed with considerable suspicion---ie, they are arguing the case for the powerful in our society and are often have a pulpit and social capital and credibility much greater than defenders of unions.

As the ongoing decline of labour in developed countries demonstrates. One side is backed into a corner and the other side must make the case that they *really* have the formula to make *everyone*'s lives better and, more importantly, the political ability to make it happen.

That's not the case, and economists shouldn't really complain when pro-union people look at them as part of the counter-revolution.

[Nick Rowe]

Mandos: sorry. But the elephant in this room right now is the thuggish behaviour of people like Jim Rootham. And that is something all of us, including people who are pro-union, should complain about, and stop.

[Mandos]

Mandos: sorry. But the elephant in this room right now is the thuggish behaviour of people like Jim Rootham. And that is something all of us, including people who are pro-union, should complain about, and stop.

But a standard behaviour of people who belong to political classes that are in a greater position of power is to police the language, including the intemperate language, of those who take positions that are lower in political power. This behaviour is also a little suspicious; unless you have reason to believe that Jim is *actually* planning to do violence to you, it seems like it is an attempt to deflect the actual point he was making.

I mean, "thuggishness"?

[liberal_usa]

"anon:"'If all prices are sticky, then an expansionary monetary policy, which increases aggregate demand, increases output and makes everyone better off.'"

me:I'm quite sure that Keynes does much the same thing. He describes his model of deficient demand, but takes fixed wages/prices into account when considering its consequences and policy responses."

That's not my reading of the General Theory. He, in fact, critiques Pigou in the appendix to Chapter 19, for assuming that wage inflexibility has a major impact on unemployment. In Chapter 19, he argues that inflexibile wages are preferable to flexible wages, and that wage flexibility would require an authoritarian government to achieve.

"In the light of these considerations I am now of the opinion that the maintenance of a stable general level of money-wages is, on a balance of considerations, the most advisable policy for a closed system; whilst the same conclusion will hold good for an open system, provided that equilibrium with the rest of the world can be secured by means of fluctuating exchanges. There are advantages in some degree of flexibility in the wages of particular industries so as to expedite transfers from those which are relatively declining to those which are relatively expanding. But the money-wage level as a whole should be maintained as stable as possible, at any rate in the short period."

or

"There is, therefore, no ground for the belief that a flexible wage policy is capable of maintaining a state of continuous full employment; — any more than for the belief than an open-market monetary policy is capable, unaided, of achieving this result. The economic system cannot be made self-adjusting along these lines."

or
"It follows, therefore, that if labour were to respond to conditions of gradually diminishing employment by offering its services at a gradually diminishing money-wage, this would not, as a rule, have the effect of reducing real wages and might even have the effect of increasing them, through its adverse influence on the volume of output. The chief result of this policy would be to cause a great instability of prices, so violent perhaps as to make business calculations futile in an economic society functioning after the manner of that in which we live. To suppose that a flexible wage policy is a right and proper adjunct of a system which on the whole is one of laissez-faire, is the opposite of the truth. "

This mean that wage inflexibility is bad, which is exactly the position that Nick is arguing the New Keynesian should not take. And Nick is right, a New Keynesian supporting unions is contradictory, as union reduce flexibility, and more flexibility is good for New Keynesians. But wage flexibility is not something Keynes argued for, which is why New Keynesians are not very Keynesian at all, as I've argued.

Take this quote from Friedman's “The Role of Monetary Policy,” American Economic Review 58, no. 1: 13.

“under any conceivable institutional arrangements, and certainly under those that now prevail in the United States, there is only a limited amount of flexibility in prices and wages.”

Sticky Prices and Wages are necessary for monetarist theories- not for Keynesian ones.

"Say's law states that there is no "general glut" in perfect competition. Yes, there might be excess demand for money, but only if the price level doesn't adjust."

I don't believe in Say's law though in cases of deficient real demand- you can't and call yourself a follower of Keynes. Otherwise, you are fundamentally a classicist of some stripe. An excess demand for money can also be called a shortfall in real demand.

"If there are unemployed resources, why doesn't their price/income fall? Why don't firms slash prices so that output rises despite depressed demand? That response would be consistent with marginal cost pricing."

I don't mean this to be snarky, but you need to read the General Theory carefully. This is where the New Keynesians and Keynes diverge. If firms slash prices, the marginal value product of labor for their goods would fall, and they would lay off workers. If you are arguing that prices, on the whole, are too high in recessions, then you are just arguing for deflation. That will not work, in fact, it will hurt. And if youi argue that firms could reduce prices and increase output, why don't they? Menu costs? Shoeleather? You'll have to do better than those.

Let's put it this way- Real potential GDP is $100 billion, but Real GDP is $90 billion and short term interest rates are at 0. Show me a series of relative price changes that will increase Real GDP to $90 billion. (Nick can join in too if he'd like). If Keynes is right (and I believe he is), you won't be able to do it. If the New Keynesian are right, you will succeed.

Relative price changes favor one sector over another, but they can't increase the total volume of production. Recessions are not (as NKs argue) about relative prices being out of whack, but, as Keynes argued, about insufficient aggregate demand. As a hint, cutting real wages will reduce aggregate demand as it decreases income. Good luck.

[liberal_usa]

"The term "monopolistic competition" may be due to Robinson. But the model of "idle" capital which I sketched above was clearly understood by J.S.Mill ("Of the Influence of Consumption on Production"), and AFAICT by Keynes as well. And it clearly relies on prices being set above marginal cost, which in modern terms is imperfect competition."

We observe idle capital due to insufficient aggregate demand and the fact the installed capital is expensive to sell as uninstalled capital. It has nothing to do with the level of competitiveness. Your argument implies that there could never be idle capital in a competitive industry, which is totally ridiculous.

Assume a totally competitive economy, with no market power anywhere. The central bank raises interest rates, reducing investment. Real GDP falls below potential, and significant unemployment results. As long as the central bank keeps interest rates too high, the price level increases more slowly, which further increases real interest rates, further reducing aggregate demand, etc. and a depression results. Prices are totally flexible. What is implausible with my story (actually my reading of the General Theory)?

If you think Keynes argued for imperfect competition being a major cause of recessions, show me the evidence. It doesn't exist- I've read the GT. Keynes always worked within a perfectly competitive framework.