Posted by
Andrews on Thursday, February 09, 2012 5:43:12 PM
I was reading
Best of the Web today and noticed a very small error, though it reveals a common, and quite large, underlying error that many people make. In the first essay of the column, Mr. Taranto argues that wages fell because of the addition of women to the work force. He does correct this slightly, but there is still an implicit error in his thinking, and if not his, in many thoughts about labor, that I want to correct.
It is one of the truisms of economics that increasing supply will, other things remaining equal, prices will fall. The reason is obvious, goods are applied to the most urgent needs first, then to the less urgent needs. As people are willing to pay less for the satisfaction of less urgently felt needs, so the highest price paid will be less for each successive use. Goods sued to produce other goods operate in a similar way. The highest bids will be for the most profitable uses, then for less profitable and so on. So prices will drop as more goods enter the market.
There are two problems treating labor in this way.
First, labor does not just add supply, but also demand. Goods generally do not. Allow me to explain. If there is $20 in an economy for burger related needs, and 10 hamburgers, burgers will sell for $2. If the number of burgers double, the quantity of demand is the same, so burgers would have to sell at $1 to clear the market. (This is not completely correct, but close enough for the example.)
Workers, on the other hand, produce goods, and those goods can be sued to exchange for other things. As Lord Say argued, goods trade against goods and supply creates its own demand. So, adding a worker alsoc reates more demand and prices of labor, that is wages, will not fall in direct proportion to the amount of labor.
The other, and bigger, issue is that labor is not a single good, or even a range of goods, it is a vast spectrum of possible goods, including considerations of training, education, commitment, hours worked and so on. Allow me to offer up a few examples.
The traditional example is the one usually used by opponents of open immigration. If you have a relatively underpopulated nation, allowing in a large number of unskilled, or moderately skilled labor will result in a net drop of wages, as the workers will compete with existing low cost labor. On the other hand, more skilled labor will probably not feel that loss, as they will not face competition. (The point being the average drops, while in many categories, there is no impact.)
However, let us imagine a very unusual situation, a nation with a largely unskilled populace, fully employed, which experiences an influx of highly skilled immigrants, bringing with them capital to invest. This would result in a net increase in average wages, and not only would the new workers experience high wages due tot he lack of competition, the demand of creating high income jobs would likely raise wages for low income workers as well.
And that is the problem with thinking of "labor" and "wages". There is no single good called labor, and no "average wages" except on government forms. There are hundreds and thousands of categories of wages which may be somewhat meaningfully compared, but even there the aggregation hides important distinctions. As I argued in "
Individual and Aggregate", lumping together data can lead to incorrect conclusions, and this is no different. There is nothing that requires an increase in labor force to decrease wages, under the right circumstances they could rise, or stay the same. Thanks to labor creating its own demand, and the many categories of labor, there is no way to easily predict what will happen when labor increases.
For example, the labor force worldwide today is many times that of 1900 or 1700 or 1500, yet we are immesasurably more wealthy with greater incomes. On the other hand, we now possess much more steel at much lower cost. In other words, economic rules do hold even over such long time spans for economic goods, but not for labor, because labor cannot be treated as a good, at least not when lumped together in an aggregate this way.
POSTSCRIPT
I realize "steel" is also invalid, as there are grades of steel, different sizes and shapes and so on, which can be treated as individual goods. Sometimes even the location of a good can make a difference (as many commodity delivery contracts make clear). Still, it is much more valid to lump together all wheat than to treat all labor as a unit. And, even if we could somehow lump together labor, there remains the fact that labor creates demand, while goods do not.