Bryan Caplan  

Cutting the Minimum Wage Really Is Good for Aggregate Demand

Is Blinder Going Off the Reser... Phelps on the Future of Capita...
Paul Krugman repeats his argument that wage cuts are an individually rational but socially destructive way to reduce unemployment.  His motive: To quash a politically impossible effort to cut the minimum wage.  Paul does address the real balance effect, but he still ignores the main arguments I've made before:

1. Cutting wages increases the quantity of labor demanded.  If labor demand is elastic, total labor income rises as a result of wage cuts. 

2. Even if labor demand is inelastic, moreover, wage cuts reduce labor income by raising employers' income.  So unless employers are unusually likely to put cash under their matresses, wage cuts still boost aggregate demand.
An even simpler way to explain it: Imagine every firm divided its existing payroll between a larger number of workers.  How is that bad for aggregate demand - or anything but good for employment?

P.S. If you prefer specific facts to textbook arguments, see Scott Sumner's legendary Table 12.2 on wages and the Great Depression.

Comments and Sharing

COMMENTS (25 to date)
Lord writes:

That is one incredibly likely "unless". More to the point, employers could already do this if they felt it in their best interest, yet they consistently do not. Do you think there is a market failure here or are they just idiots?

Tom West writes:

1. Cutting wages increases the quantity of labor demanded. If labor demand is elastic, total labor income rises as a result of wage cuts.

How so? If we halve wages and increase labor demand by 50%, I calculate 0.5 * 1.5 = 0.75 the previous total labor income.

2. Even if labor demand is inelastic, moreover, wage cuts reduce labor income by raising employers' income. So unless employers are unusually likely to put cash under their matresses, wage cuts still boost aggregate demand.

Between employer and laborer, wouldn't the total amount of income remains constant? In terms of how the income is spent, I suspect, but have no proof, that lower wage earners are more likely to spend money on local products like rent, food, etc.

ryan yin writes:

Tom West,
(1) If wages dropping by half leads to only 50% more labor demanded, then by definition we are not in the elastic demand case. (2) The income isn't constant unless we're in the perfectly inelastic case. It's possible that employers don't consume as much as employers, but the key question is whether they stick it under their mattresses. (Income invested doesn't count as a leakage.)

If employers want to cut wages below the minimum wage and if such a wage would in fact be an equilibrium, then the reason they don't can't possibly have anything to do with market failure.

david writes:

@ryan yin

Income saved does count as a leakage when interest rates are zero. It sits under a bank's mattress instead of your mattress, perhaps.

ryan yin writes:

It's possible that all profits that aren't eaten sit under a mattress, but I'm not sure that's the case anymore. The amount of investment is nonzero.

Marlo writes:

Can someone enlighten me about point (2)? I read it as "taking money from workers and giving it to employers raises aggregate demand". By that logic, wouldn't wages that get lower and lower always be best?

roh writes:

Krugman says that "if a subset of the work force accepts wage cuts, it can gain jobs." He then argues that if all workers take a pay cut, this no longer applies. However, last time I checked, only a subset of the workforce earns minimum wage. His argument that cutting all wages won't increase employment doesn't mean that cutting the minimum wage won't increase employment. He seems a bit disingenuous.

Tom West writes:

Ryan Yin, can you clarify? I'm not an economist, but I thought the definition for elastic is roughly the same as Wikipedia's:

In economics, elastic describes demand that is very sensitive to a change in price.

So, halving wages to increase more labor hours worked would still indicate the labor market is elastic.

Likewise, I thought inelastic would mean that the same number of hours were worked regardless of the wage.

Is there a different economics definition of elastic/inelastic that I'm not aware of?

Lord writes:

I was referring to wages in general. As to the minimum wage there are so few earning it, it would not be a large effect. If Bryan really likes this he should support reducing the work week which would be much more effective with the same result. Course there are a lot of complications, like workers both are and have both fixed and variable costs.

Redland Jack writes:

@ Tom

If a 1% change in prices would result in a less than 1% change in quantity demanded, it's inelastice. (Note: the 1% should really be the 'marginal' effect, or the derivative). If you are in an inelastic point in the curve, raising prices (a small amount) increases revenue. The exact opposite is true if you are in the elastic portion. (Prices and quantity demanded are the usual way I think about it. You can just swap in wages and labor demanded).

Locale writes:

Tom West,

Elastic demand basically means that as price changes, the x-axis will respond heavily (can be negatively or positively). Inelastic would suggest that for a price change, the market wouldn't respond. On a graph, inelastic demand would have a steeper line (curve) than that of elastic.

There are some calculations to determine how elastic/inelastic demand are between two points on a curve. The example of half is more drastic than say 1/7th or 1/4th less.

In claiming that wages and labor are elastic especially in the short run, we can associate this to a market that will respond to small changes more readily. Inelastic markets don't respond quickly since items such as equipment, building space, and time factor in. Those take time to build and manufacture.

Redland Jack writes:

@ Marlo

The company would certainly be producing more output. Under reasonable assumptions about the price of the output (that a 1% increase in output results in a

I think... if I screwed this or my last comment up, I'm sure someone will point it out!

CJ Smith writes:

@ Tom West:

In economics, elastic demand means that an X% decrease in price causes a more than X% increase in units demanded; thus, total sales revenue increases. A unit elastic demand means that an X% decrease in price causes an X% increase in units demanded; thus, sales revenue remains the same. Inelastic demand means that an X% decrease in price causes a less than X% increase in units demanded; thus, total sales revenue decreases. Perfectly elastic demand means that any decrease in price causes demand to become infinite, any increase cause the demand to drop to zero. Perfectly inelastic means that price has no effect on demand - the same amount is demanded regardless of price. All elasticities should be considered within bounded price ranges, but are frequently considered unbounded for simplicities sake.


How to you come up the conclusion that halving wages and doubling consumers would solely increase aggregate demand absent a predominantly demand inelastic aggregate economy? Instead the demand would shift away from traditionally price elastic goods to traditionally price inelastic goods, which would cause a shift in aggregate supply curves as suppliers exited formerly profitable markets. I would think, at best, the effect would be indeterminate.

Burk writes:

Sounds like the ideal situation would be that workers get paid nothing, and employers collect all the income. Would that lead to an ideal amount of aggregate demand? As I recall, that was the situation in late Roman times (with slaves), among many others, and it didn't turn out so well.

Nothing like a bit of ideology to guide one's "economics"!

Doc Merlin writes:

Wow, you totally didn't read that right, Burk.
He is talking about removing a PRICE FLOOR, not forcing people to pay less.

mark writes:

Although usually aligned on the Econlog side of debates with Krugman, here I find myself on the other side. The theory is fine but the real world doesn't produce anything approaching a linear tradeoff. If you cut wages to nurses in half, you won't get twice as many nurses. You might even have a net loss of nurses and patient welfare probably suffers in most scenarios. If you take an unskilled labor force like a fast food restaurant, there are factors imposed by the physical layout behind the counter that constrain how many people you can add. There are constraints in numbers of hours an employment location will be open. Skilled workers often have bargaining power that deter employers from risking a cut in their wages. Etc. I don't think this is the optimum approach in a modern economy. I'd rather govt invest in infrastrucure improvement to create jobs and enhance welfare.

David C writes:

I think Paul Krugman has come to the conclusion that we need inflation by any means necessary. So he's opposed to wage cuts because they'll lead to deflation.

Tom West writes:

Thanks to all who answered. Doing some googling, you can find both definitions as above. has Redland's Jack and CJ Smith's definition (although it seems a bit careless. Does 100% decrease in price bring about only a 100% increase in demand? I suspect they really mean elastic goods are above the P * D = Constant curve.)

Wikianswers and Wikipedia use the more vague definition, which is called "relatively elastic" by

Anyway, thanks to all who answered. I learned some more economics today!

Eric H writes:

How are wage cuts different from job sharing?

Eric H writes:

My question was in regards to this sentence:

"An even simpler way to explain it: Imagine every firm divided its existing payroll between a larger number of workers."

simmmo writes:

rajiv sethi takes caplan's argument down.

toddorbert writes:

"Imagine every firm divided its existing payroll between a larger number of workers. How is that bad for aggregate demand - or anything but good for employment?"

Could someone explain in layman terms the logic behind the assumption that when the wages of a person who is working on minimum wages are decreased, he does NOT decide to work more to compensate the losses in income? Because if he does, that means that the employment measured in number of persons employed surely will decrease due to decrease in minimum wages?

Bill Woolsey writes:

Some of the comments here appear to ignore supply.

If there is a surplus of labor, quantity demanded is less than quantity supplied. The actual quantity is on the short side of the market--quantity demanded.

A lower wage increases quantity demanded and reduces quantity supplied. Because actual quantity is on the short side of the market, it expands. Lower wages increases employment.

If there is not a surplus, and quantity demanded equals quantity supplied, and wages go down, then there will be a shortage. Quantity supplied will be less than quantity demanded. And so there will be less employment.

There is no "lower wages always good" or "higher wages always good." Wages should be at a level where quantity demanded equals quantity supplied.

If we are really talking about the mimimum wage, then the macro-effects are trivial. The income earned by mimimum wage earners is tiny. The share of total costs representing minimum wage labor is tiny. It is like 2% of workers make the mimimum wage. But remember, the total income of the lowest 20% of income earners is only about 4% to total income. The ceteris paribus macro impact on aggregate income of a 10% cut in the mimimum wage is a tiny fraction of 1%.

Caplan was doing some kind of analysis where we assume all wages drop in proportion, which would have major macro impacts. Wages are 66% of total income. If all of them fall 10%, that is about 7% less income. And what happens to profit and other capital income, what happens to the incomes of workers hired, and so on, is important.

I think the best evidence is that a lower mimimum wage would result in more unskilled workers having jobs, but the total income of all of them together would be less, and this would be an insigificant impact on total income. The prices of the products of unskilled labor would be a bit less, and people would buy a bit more (which raises the demand for those products and is the reason there are more jobs for unskilled workers.) That people spend less on those products at lower prices leaves them more money to buy other things. But this is a tiny, tiny impact.

A partial, micro analysis is the best way to understand changes in the minimum wage. The macro analyais of a proportional change in wages, on the other hand, is not usefully understood using micro tools. It is all about changes in real money balances, the pigou effect on consumption and saving, and the like.

Redland Jack writes:

@ Tom


A quick point of clarification. The elasticity is at a specific point/intersection. That is, if you are in the elastic point of a curve, a .1% change in price will generate a >.1% change in quantity demanded. However, at some point, if you keep changing the price, you'll move from elastic to inelastic.

The main thing, is that (unless specially constructed), the curves will always be elastic on one part of the curve and inelastic on the other.

@ burk

A lot of economists aren't particularly concerned with aggregate demand, one way or the other. AD is, largely, a Keynesian concern. Also, as Bill notes, you'd have to look at labor supply as well as quantity demanded (though as you point out, with slavery, supply is, I suppose relatively fixed).

Anyhow, even the Keynesians wouldn't (I don't think) want AD to increase to infinity. They just want it to go to the highest equilibrium point, as opposed to settling at an equilibrium that doesn't maximize GDP.

luis writes:

There was no minimum wage in the early 1980s or 90's in Uk , where I live and employment , including and especially low wage employment tanked . Companies were entirely free to offer lower wages which workers were entirely free to accept . They chose freely not to do so . A minimum wage was introduced in 1999 and is now about double in real terms what it is in the US and until relatively recently low wage job growth has soared . Do you think that if CafeHayek had a shred , slither , morsel , titbit , crumb or iota of evidence to suggest job loss from this WE WOULDN'T KNOW ?!!!!!You may rest assured that the equivalent people here have been working round the clock to find any such evidence and failed to do so. EVEN THE ECONOMIST , which hails Adam Smith and Milton Friedman as the two greatest ever economists has admitted "We remain opposed to the minimum wage as a matter of principle but are nonetheless forced to accept that the evidence is that it has had no significant effects on employment " We were told that it would cost2 million jobs . Not a bad guess . They were only out by 2 million . I admire the faith of the Friedmanites because faith it most certainly is.

Comments for this entry have been closed
Return to top