≡ Menu

Inequality IV

In the earlier posts in this series, I talked about what we mean by opportunity and mobility and why various measures of mobility, inequality and absolute poverty might be distorted by various changes unrelated to opportunity.

In the last post, I gave the example of divorce and family structure.  Between 1970 and 2000, there was a large increase in female-headed households which distorts any interpretation of a snapshot in 1970 and 2000.

The other major change is immigration.  Here are the numbers on the proportion of the workforce that is foreign born:

1960     6
1970     5.2
1980     6.5
1990     8.8
2000     13.2

(These numbers are taken from a paper by Borjas and Katz that focuses on Mexican immigration but these numbers are for all immigrants.  The paper only has a chart but I’ve received the actual data from the authors.  These numbers are not easy to find, btw.  Borjas and Katz created them with a great deal of effort by digging into very large Census databases.)

The surge in immigration brought workers with less education and lower wages than Americans who were already here.  As the proportion of workers who are foreign-born increases, that means that the average wage will be pushed downwards. 

That does not mean that the average person in 1970 is falling further and further behind.  The example my co-host here at the Cafe, Don Boudreaux uses, is that when Shaquille O’Neill enters a room, the average height rises but no one got any taller.  Similarly, immigrants with low skills will drive down the measured average wage even if their arrival has no impact on those who are already here.

Unlike the Shaquille O’Neill example, immigrants will affect the wages of people who are already here, increasing some people’s wages and lowering others.  There’s a big debate about  the net impact which I’m going to ignore here.  My point is that independent of those complex effects which are quite difficult to measure, there is a pure distorting measurement effect or composition effect if you naively compare two snapshots over time and make conclusions without correcting for the dramatic change in the composition of the workforce.  It’s not the same people.

Having said that, I’m skeptical that going from 5% to 13% is going to have that big an effect on the overall average.  That would depend on the ratio of immigrant wages to native-born wages.  But the point here is that when you look at immigration and the change in female-headed households over the same time period that I wrote about in the last post, you get a feel for how hard it is to compare two snapshots over time if they’re not the same people.

Which brings us full circle to the Wall Street Journal article of last week that began this series of posts.  I quoted the first part of the headline in that first post.  Here’s the whole thing:

As Rich-Poor Gap
Widens in the U.S.,
Class Mobility Stalls

Those in Bottum Rung Enjoy
Better Odds in Europe;

How Parents Confer an Edge

Immigrants See Fast Advance

Immigrants See fast advance?  How can immigrants advance rapidly if class mobility is stalled?  Something is wrong with this picture.  In fact, the surge in immigration suggests that class mobility is very much alive and well in the United States.  People very much prefer to be poor in the United States than poor elsewhere.  Not just because poor people are better off here than in other countries.  Unfortunately, the article only has two sentences to justify the sub-head about immigrants:

One drawback of the surveys is that they don’t
capture the experiences of recent immigrants or their children, many of
whom have seen extraordinary upward mobility. The University of
California at Berkeley, for instance, says 52% of last year’s
undergraduates had two parents who weren’t born in the U.S., and that’s
not counting the relatively few students whose families live abroad.

But immigrants do make rapid progess.  I hope to have some data for you on that tomorrow and close out this series with some empirical findings that follow the same people rather than just use two snapshots.