Immigration Lowers Wages. What About Prices?

One of the arguments commonly bandied about in the immigration debate is that immigration lowers wages. Paul Krugman writes,

First, the net benefits to the U.S. economy from immigration, aside from the large gains to the immigrants themselves, are small. Realistic estimates suggest that immigration since 1980 has raised the total income of native-born Americans by no more than a fraction of 1 percent.

Second, while immigration may have raised overall income slightly, many of the worst-off native-born Americans are hurt by immigration — especially immigration from Mexico. Because Mexican immigrants have much less education than the average U.S. worker, they increase the supply of less-skilled labor, driving down the wages of the worst-paid Americans. The most authoritative recent study of this effect, by George Borjas and Lawrence Katz of Harvard, estimates that U.S. high school dropouts would earn as much as 8 percent more if it weren't for Mexican immigration.

Okay. But isn't there another side to this phenomenon, namely prices?

If immigrants are willing to work for lower wages for one company, other companies will similarly bid for their labor at slightly higher but still low wages. These savings will likely be reflected in price reductions given a sufficiently competitive market and elastic demand.

In an economy in which wages hypothetically decrease 10% but prices decrease 15%, aren't people richer? Granted, this is a simplistic statement that doesn't go into which people are richer and which ones save on prices, nor into the relevant time period, but the effect of immigration on prices shouldn't be ignored.

Share this

It's just Wal-Mart-ophobia

It's just Wal-Mart-ophobia all over again -- seeing only the supposed wage effects and conveniently ignore the benefits of low prices for all.

It's not only low prices,

It's not only low prices, there's also the type of services that evaporate at higher labour costs. Some old geezer who can get his lawn mowed at a certain rate will probably just let it go to seed if restrictions raise that rate. He's down a tidy lawn and some Mexican guy is down the wages he would have earned for cutting the lawn. That's a lose-lose.

I was listening to the Rush

I was listening to the Rush Limbaugh show yesterday (I know, I know ...), and a caller happened to be a neighbor of Mr. Limbaugh's on Palm Beach. The caller owned an agricultural company in Florida, and he all but admitted that he employed illegals, as I am sure most or all such companies around Florida do.

He kept saying the same thing over and over again -- "American citizens will not take these jobs." Not even the legal immigrants will take them, because they can get better jobs elsewhere.

What the caller did not say (but what he really meant, of course) was that no one but illegals will take jobs with his company at the wage rate he is offering.

I am sure that if someone suggested that he had to raise his wage rates in order to hire citizens or legal immigrants, his response would be that he could not afford it, that he would then have to raise the prices of his products, which he could not do because he is in competition with other companies who hire illegals, that they would underbid his prices, and he would thus be out of business.

Krugman's position, I am sure, would be that the owner of the company ought not to be living on Palm Beach, that his accumulated wealth should be liquidated for the benefit of the proletar... I mean, his employees.

The Austrian position, however, is that a market-based wage rate should not be tampered with, that immigration restrictions are a classic form of protectionism for domestic workers, that they obviously raise prices of the products in question (if these laws are enforced), costs which are borne by everyone, and the restrictions don't even accomplish their stated goals anyway and lead directly to the creation of a black market.

Hmm, I'm more of a

Hmm, I'm more of a businessman than an economist but I'm hard pressed to think of a circumstance where lowering wage costs by 10% would allow me to lower my product or service cost by 15%. Even in the best case with a zero overhead 100% service business I could only see a 10% reduction.

Can you explain how you arrived at the 10% and 15% figures?

Keving- Walk with me for a


Walk with me for a bit o the hypothetical-

If all the companies of type A, say a goods manufacturer or supplier, reduce labor costs by 10%, then assuming competition grinds profit down to 'normal', you'll get a 10% reduction in prices for the stuff provided by Companies Type A.

Company Type B is a different kind of supplier that buys things from Type A's. Its now getting Type A goods for 10% off. Assuming the same "competition grinds away entrepreneurial profit", you'll get a price drop for Type B's goods.

If Type A buys stuff from Type B, then you get a recursive price dropping where the end result is some amount greater than the original cut. If you postulate a C, D, E, etc, type companies, all of whom are downstream from the initial cost reduction, you'll also get a similar price dropping that, in total, is more than the original cut.

Thats a simple proof of possibility. You can think of it similarly in the way of fractional reserve banking. Ultimately the total amount of new cash injected into the system from a starting point is the reciprocal of the reserve fraction- If you have 90% reserves, you get $1.11(bar) dollars created for every dollar injected. If you have 10% reserves, you get $10 for every $1 injected.

Now, not every company in the hypothetical is going to have the same mix of labor/good costs, but assuming every company in the downstream chain has 33% of its costs come from buying goods from the company upstream, then a 10% cost reduction at the beginning will give you a 15% reduction in prices overall. (1-0.33 -> 0.67 -> 1/0.67 = 1.5x multiplier)

IN the real world it is, of course, not that simple nor that mechanical, but the principle is the same, and it works the same in reverse.

Oyvay. That's all very

Oyvay. That's all very clever except that if you reduce your labour costs by 10% you don't get a 10% reduction in you product or service cost -- unless you have no debts, depreciation, taxes rent etc.

If you have a hypothetical company in which labour comprises 25% of their cost base then a 10% reduction labour cost produces a 2.5% reduction in price if all the savings are passed through.

I don't know a reserve fraction from a 2 by 4 but I do arithmatic.

Keving- Good point, lets


Good point, lets look at that-

Keeping the original heroic assumptions, and then assuming that the first company in the chain also has a 1:2 ratio of nonlabor:labor costs as the ones later in the chain, that would give an initial price cut of only 6.7% for a 10% cut in labor costs.

The price cut translated over the infinite number of folk in the chain = 10.05% price reduction. Still a slightly positive deal even for the laborers in the start of the chain, their net/real wage has gone up. Every other worker's real wage increased by 10%.

Let's flip it around, as you do, to labor only being 25% of the costs. The initial price drop is as you say only 2.5%. However, assuming every company has the same ratio of costs, the multiplier shifts WAAAAY in our favor- 75% of every downstream company's costs are now reduced by 2.5%, instead of only 33% of costs cut by 6.75%. Interestingly enough, the 2.5% cost cut in the initial company in the chain gives us precisely 10% in lowered cost total. So the initial workers are even money and the rest of the economy is up again by 10%.

Also, in all of this, it is

Also, in all of this, it is assumed that each company buys just from the next up in the chain, but that every laborer/individual buys a basket of goods of equal proportion to every type of company. Nick, Patri, if you guys know some nifty math that can help elucidate this better, by all means... :)

Brian- First. Where do you


First. Where do you get this: "75% of every downstream company’s costs are now reduced by 2.5%"? In what way is the number of downstream companies dependant on the labor cost base? Increasing or decreasing my labor cost as a percentage of my total cost does not affect the number of people who directly or indirectly consume my product or service.

It seems to me like you take a money supply theory and are trying to shoehorn it into a circumstance where it doesn't work -- if only because services don't flow through the economy in the same way dollars do.

I agree with your main point that reducing labour will reduce cost but it seems to me you should have stopped before the last paragraph.

I also agree that you've ignored who is getting rich and who is getting lower wages but you have still drawn the conclusion that "people are richer" by assuming the people with reduced wages are reaping 100% of reduced costs.

So that works in reverse eh?

So that works in reverse eh? If that's the case then it should also be true that prices should go up by a greater percentage than an increase wages. Let's test that then:

I'm a farmer who has hired a Mexican for $5/hour. I give him some nunchucks which double his output to $20 worth of corn per hour. The nunchucks cost $10 and only last a 40 hour work week. So I currently have an $800 income minus $210 in expenses ($200 in labor plus $10 for nunchucks) for a week. That gives me a profit of $590 which is a 75% profit margin.

Then one day it's decreed that my Mexican must earn 60% more, but I'd like to keep my 75% margin. That means I have to increase the price of my corn, giving me the following equation: IncomeCorn - (Labor + 60% * Labor) - Nunchucks = ProfitMargin * IncomeCorn

Working that out, I find the income from the corn needs to be $1320. That turns out to be a 65% increase from before which is inline with what you said.

Now if the nunchucks were $100 instead of $10, then this is a different story. Before the 60% wage increase, I would have made $800 minus $300 in expenses. The $500 in profit gives me a margin of 62.5%.

With the wage increase of 60%, I find that the income from corn in a week needs be $1120 to preserve my margin. That turns out to be a 40% increase in price.

That's counter to your claim. Even if the nunchuck supplier had to increase his prices by 40%, my increase in price would still only be 53%.

And it's beyond my ability to do the math to figure out the increase if the nunchuck supplier also eats corn...

Keving- In order: First.


In order:

First. Where do you get this: “75% of every downstream company’s costs are now reduced by 2.5%"? In what way is the number of downstream companies dependant on the labor cost base? Increasing or decreasing my labor cost as a percentage of my total cost does not affect the number of people who directly or indirectly consume my product or service.

Go back to my initial example if you want to know how the chain works. But in a nutshell (again), assuming a series of companies A through Z that buy one from the other in series, and that the change in labor cost is only to company A, then first the labor cost is modified by the proportion of labor in A, which affects the price of A's output which is bought by B- this price reduction affects B's inputs, not labor. So if you postulate (as you did in your counter-hypothetical response, remember) that every company has 25% labor costs and 75% input costs then the 10% wage cut in A results in 2.5% cheaper goods to B. 75% of that 2.5% cut is reflected in B's prices, and so on and so forth until it becomes infinitesimal. Added all together it is equivalent to multiplying the initial cost reduction by the reciprocal of affected proportion of costs.

But none of that is necessary- you can posit a 10% reduction in wages across the board, without relying on serial transfer. The result then is even more clear- reghardless of the mix of labor to non-labor costs, the total price reduction is always equal to the wage cut, so there is no change in welfare. But since we're only talking about wage effects on one sector of the economy (low-skilled), it would not be appropriate to model an across-the-board wage drop. But that's neither here nor there.

The point of the simple model is to prove the possibility of the concept. Obviously there are going to be limits to how much downstream producers will lower prices versus change production mixes (more, less), demanding more labor, etc., and how quickly cost savings can percolate through (speed of competition), etc, etc, etc. The proper response, then, is to figure out a hypothetical where the reduction in costs results in pure profit to company A and is never passed on to anyone else via competition; in any other case where cost savings are passed along, the workers of the economy as a whole will be better off (even if the ones in sector A are not, relatively speaking).

It seems to me like you take a money supply theory and are trying to shoehorn it into a circumstance where it doesn’t work – if only because services don’t flow through the economy in the same way dollars do.

This is silly. Of course they do. Dollars are exchanged for (wait for it) *goods and services*. For every dollar that moves via exchange through the economy an equal and opposite amount of goods and/or services moves the other direction. Its a logical impossibility for it to be otherwise. Thus any aggregation of the economy will work *precisely* on the mathematical (not monetary) principle of infinitely summed fractions, so long as (a) there is transfer from one entity to another and (b) some bit is passed along on net, from one to another [that is, A gets something from B, has something left over and thus passes it on to C].

I also agree that you’ve ignored who is getting rich and who is getting lower wages but you have still drawn the conclusion that “people are richer” by assuming the people with reduced wages are reaping 100% of reduced costs.

Yes, the moderl *assumes* that profit margins and wages for everyone else remain the same, aside from the sector affected by the low wage competition. And by nature of the model *of course* we've assumed that everyone benefits from the reduced costs. If you restrict the subset of goods that consumers (the sum of all workers in all sectors) buy to goods early in the chain then the Subset(A) workers will be adversely affected so long as you exclude enough downstream company's products that the cost savings don't add up to the wage reduction. But in any scenario the remainder Consumers(All minus A) will be proportionally better off, since their wages are not dropping but they're getting cheaper goods.

Neither objection challenges the conclusion that economy-wide and for most workers the cost reduction is a major boon.

Nolan- Your argument


Your argument stumbles from the starting block, in that you do not get to *choose* your profit margin; by and large (or one might say by definition) your profit margin is determined by the intersection of demand and your costs. Your ability to profit is limited in absolute terms by the willingness of people to pay. Assuming you are in a competitive environment prior to the cost increase, ANY increase in cost to you will reduce your profit so long as demand curves slope downward.b It is for this reason that my little explanatory model works as it does- costs go down, profits go up initially but in a competitive system there are incentives to cheat (lower prices a bit and reap extra profit at the expense of your competitor) such that tit-for-tat competition on price will reduce the margin back to where it was an equilibrium before (i.e. the normal rate of return on capital).

Hence the rest of your comment is rendered moot.

Brian- I understand how your


I understand how your chain works.

When you wrote "75% of every downstream company’s costs are now reduced by 2.5%" I read 75% of all the downstream companies have their costs reduced by 2.5% when you meant for each of the downstream companies they have a 2.5% reduction applied to 75% of their cost base.

I don't think your model is a very good representation of the actual situation and I remain unconvinced that the model provides any useful results given all the assumptions. For a start you've ignored all products and services which are chain terminals. Reducing the price of cutting my lawn ends with me. I don't pass on any cheaper goods as a result of my cheaper yard maintenance or haircut. Since the vast majority of low wage work provides these types of services, does the series math really hold up?

Keving- In all of your


In all of your examples of dead-end services (no pass through), there is absolutely no negative to you and only gain from lower wage substitutes. After all, employing a mexican to cut your lawn doesn't lower your wage. And while yes he's doing something that a native (you) could do, you don't want to do that job- your time is worth more to you elsewhere than the wage you pay the mexican, and as a result you (a) have more time free to do other, higher valued activities, (b) you get a cut & tidy lawn, and (c) Jose gets money that he values more than not doing work.

So I'm not sure why you've retreated to end-user examples because those are *absolutely* in favor of low wage work.

But further, you *do* pass along the savings. You either spend the money saved on something else (increasing economic activity, helping employ someone else), or you save the money (by taking it out of the money supply by burying it [lowering demand and raising the value of all remaining cash], or by giving it to someone else to use in exchange for interest). That is a boon to the rest of the economy, directly attributable to the reduction in cost to you.

Frankly, it doesn't matter where the chain ends or how. But let's get the worst case scenario out of the way-

There is no chain. No other producer uses anything from Sector A, Sector A is strictly consumption goods. Wages are dropped 10% due to oversupply of workers in sector A (a low or no-skill sector). If labor costs are only 25% of the prices of goods from sector A, then sector A laborers are hit hard with the 10% loss and only a tiny reduction in prices (2.5% times whatever share of the total market Sector A has, since we stipulate that every consumer spreads their consumption equally across all sectors' production). *For all other sectors, their real wage increases by the reduction in price for Sector A goods.*

In *any and all* scenarios, reducing the costs of an economy will make the aggregate economy better off. In almost every scenario were ceteris is paribus, even the poor folk in the wage depressed sector are better off via price offsets.

Thus immigration's effects *per se* are neither broadly or generally negative nor are they a basis for rejecting low wage immigration.

Saying that wages would go

Saying that wages would go up 8% if not for illegal immigrants doesn't mean that illegal immigrants lower the cost of labor by only 8%. If all of the illegal immigrants left, a lot of the jobs currently held by illegal immigrants wouldn't go to dropouts; they would simply disappear. Companies that rely heavily on immigrant labor would reduce production or go out of business entirely. Many industries would be radically transformed. Krugman is saying he thinks the average wage of dropouts would be 8% higher in that completely different world where we don't have illegal immigrants. That's very different from asking how much individual firms save by hiring illegal immigrants as opposed to what they would pay for dropouts without illegal immigration.

All of you are assuming that the increase in dropout wages if illegal immigration were stopped is equal to the labor savings that companies enjoy by hiring illegal immigration. I don't think that's true.

Brian- I haven't argued


I haven't argued against lower wages. I'm relatively wealthy and I can don't dwell overly much on the price other people pay for my cheap good.

I haven't, as far as I know, retreated anywhere. I started out trying to undertand how you thought a 10% decrease in wage cost produced a 15% final product cost reduction. You explained that you created a model with a bunch of assumtions -- in particular that there was a chain of other producers serially benefitting from the wage decrease. I then pointed out that this model and it's assumption of serial producers is flawed.

To justify your model and it's series summation it is necessary that a series exists. It doesn't seem to me that it does.

You've made a cogent argument but only so long as all your assumptions are valid. I don't think they are.

The original point was not

The original point was not that a 10% reduction in wages will necessarily result in a 15% drop in prices, but that focusing on the nominal wage reduction while ignoring the price reduction overstates the degree to which immigration harms certain native Americans.

About Krugman's

About Krugman's comment...

You have to sense disingenuousness when an argument against illegal immigrants is so that we can show that we care about high-school dropouts. Not that we don't want them to dropout; I guess we need them because us educated rich folk always want an underclass to underpay.

Brian: So I don't get to

Brian: So I don't get to choose my margin, but eventually it would normalize to what it was before? If that's the case, then the rest of my comment should apply. If not, then we won't see an immediate reduction in prices since a producer's margin will have gone up with reduced wages.

It's also likely that prices would stay the same too, ie: increased demand from working immigrants.

What nonsense. The costs to

What nonsense. The costs to lower waged workers would drop by 10% ( and not 15%) only if

a) Every wage earner - not just the unskilled - reduced their wages.
b) if the savings in wages were passed onto the wage earner rather than being realized as greater profits. That is not the case in the US now.
c) If the US did not import goods from anywhere.
d) If immigrants did not put upward pressures on rents and housing
e) If the US was not a social welfare state. Every low paid worker is a drain on the economy. They do not pay enough for the sum total of their children's education and their social welfare entitlements, nor the pension that will accrue to them ( I mean unskilled workers in general - a population you do not want to increase - illegal immigrants are a greater drain as non-taxpayers).

Most of your counter arguments are just dumb. Reduce the working population by 8% and companies will be forced to mechanise - which is the only historical process for increasing productivity per employee, and wage per employee can only increase with per-capita increases in productivity.

If we are to argue the "moral" argument that making people poorer makes them richer then it seems the 19th century must have been richer than now, and liberia has to be reconsidered as a economic paradise. In the real world when people get poorer they get poorer. When they earn less, they get poorer. If they continue to get poorer then they will be unable to afford many goods - for instance cars - which can be produced at a certain prices. These are dependent on input costs which are independent of wage prices, and even if wages were zero, the costs would be high enough. Some services may get cheaper, but not all ( not the services protected by elite workers who do not have to compete, or compete less, with immigrants) /

Unskilled services are cheap in the Third World now. The Third World isnt as poor as it seems from the raw statistivs as services are way cheaper there: however goods are not on the whole cheaper and earning $1 a week may get you a haircut for 20C (equivalent) but it wont buy you an iPod.

please get rid of the dancing smilies. they eat CPU.

"b) if the savings in wages

"b) if the savings in wages were passed onto the wage earner rather than being realized as greater profits. That is not the case in the US now."

should have been "passed onto the consumer", of course.

"It’s not only low prices,

"It’s not only low prices, there’s also the type of services that evaporate at higher labour costs"

the ending of child labour ended any number of services, people started to clean their own chimneys rather than stick a ten year old up a soot and tar infested pipe. The arguments in favour of lower wages, immigration, and so on are the very same as the arguments in favour of retaining child labour ( jobs that adults would not do, factories would collapse), and even slavery ( which was very good for the economy of the slave owning states, if not the actual slaves.).

"The arguments in favour of

"The arguments in favour of lower wages, immigration, and so on are the very same as the arguments in favour of retaining child labour ( jobs that adults would not do, factories would collapse), and even slavery ( which was very good for the economy of the slave owning states, if not the actual slaves.)."

Not at all. We argue in favor of immigration because we suspect it will take place. Child labor, on the other hand, we do not honestly suspect people will engage in it. Children are young and unskilled, they fetch paltry wages. So, parents get minimal financial reward by putting their children to work, while at the same time they give up on school for their children, sacrificing greater wages later on.

You must remember, child labor was on the decline long before it was outlawed.

As for Slavery, WTF? We are libertarians, what makes you think we would be in favor of the government forcing people to work as slaves? Slavery cannot exist without a "state" of some-sort to enforce it, otherwise, what is to stop me from running away?

So, do away with all the laws you mentioned, and immigration will increase, a good, Child Labor will still almost never take place, and Slavery, if it existed, would soon evaporate, unable to persist without the iron fist of the state.

Minimum wage level labor is

Minimum wage level labor is only a tiny contribution to overall prices. After all, most manufacutring is done overseas where the labor costs only $1/hour.

There are only a few things where there would be a noticeable rise in prices, like lawn care, but really we'd survive as a nation if people's lawns weren't manicured quite as nicely.

We'd survive as a nation if

We'd survive as a nation if we all cut off our left arms just above their respective elbows. It's still a lousy idea.

^ "I must concur with the

^ "I must concur with the good doctor on this issue"

What about the benefit to

What about the benefit to illegals themselves? Surely the gain more in increased wages than low income natives lose in reduced wages. There is a net benefit here even before taking into account price effects. Much the debate around immigration doesn't make much sense unless you don't value benefits to brown people who happen to be born on the other side of some arbitrary geopolitical boundry.