No subject

Fri Feb 24 11:55:10 PST 2006

Do immigrants really hurt American workers' wages?

Get article background

EVERY now and again America, a nation largely made up of immigrants and
their descendants, is gripped by a furious political row over whether and
how it should stem the flood of people wanting to enter the country. It is
in the midst of just such a quarrel now. Congress is contemplating the
erection of a wall along stretches of the Mexican border and a crackdown on
illegal workers, as well as softer policies such as a guest-worker programme
for illegal immigrants. Some of the arguments are plain silly. Immigration's
defenders claim that foreigners come to do jobs that Americans won't‹as if
cities with few immigrants had no gardeners. Its opponents say that
immigrants steal American jobs‹succumbing to the fallacy that there are only
a fixed number of jobs to go around.

One common argument, though not silly, is often overstated: that immigration
pushes down American workers' wages, especially among high-school dropouts.
It isn't hard to see why this might be. Over the past 25 years American
incomes have become less equally distributed, typical wages have grown
surprisingly slowly for such a healthy economy and the real wages of the
least skilled have actually fallen. It is plausible that immigration is at
least partly to blame, especially because recent arrivals have
disproportionately poor skills. In the 2000 census immigrants made up 13% of
America's pool of workers, but 28% of those without a high-school education
and over half of those with eight years' schooling or less.

In fact, the relationship between immigration and wages is not clear-cut,
even in theory. That is because wages depend on the supply of capital as
well as labour. Alone, an influx of immigrants raises the supply of workers
and hence reduces wages. But cheaper labour increases the potential return
to employers of building new factories or opening new valet-parking
companies. In so doing, they create extra demand for workers. Once capital
has fully adjusted, the final impact on overall wages should be a wash, as
long as the immigrants have not changed the productivity of the workforce as
a whole.

However, even if wages do not change on average, immigration can still shift
the relative pay of workers of different types. A large inflow of
low-skilled people could push down the relative wages of low-skilled
natives, assuming that they compete for the same jobs. On the other hand, if
the immigrants had complementary skills, natives would be relatively better
off. To gauge the full effect of immigration on wages, therefore, you need
to know how quickly capital adjusts and how far the newcomers are
substitutes for local workers.

City to city

Empirical evidence* is as inconclusive as the theory. One method is to
compare wage trends in cities with lots of immigrants, such as Los Angeles,
with those in places with only a few, such as Indianapolis. If immigration
had a big effect on relative pay, you would expect this to be reflected in
differences between cities' wage trends. David Card, of the University of
California, Berkeley, is one of the leading advocates of this approach. His
research suggests that although there are big differences between cities'
proportions of immigrants, this has had no significant effect on unskilled
workers' pay. Not everyone is convinced by Mr Card's technique. His critics
argue that the geographical distribution of immigrants is not random.
Perhaps low-skilled natives leave cities with lots of immigrants rather than
compete with them for jobs, so that immigration indirectly pushes up the
supply of low-skilled workers elsewhere (and pushes down their wages). Mr
Card has tested the idea that immigration displaces low-skilled natives and
found scant evidence that it does.

An alternative approach, pioneered by George Borjas, of Harvard University,
is to tease out the effect of immigration from national wage statistics. Mr
Borjas divides people into categories, according to their education and work
experience. He assumes that workers of different types are not easily
substitutable for each other, but that immigrants and natives within each
category are. By comparing wage trends in categories with lots of immigrants
against those in groups with only a few, he derives an estimate of
immigration's effect. His headline conclusion is that, between 1980 and
2000, immigration caused average wages to be some 3% lower than they would
otherwise have been. Wages for high-school drop-outs were dragged down by
around 8%.

Immigration's critics therefore count Mr Borjas as an ally. But hold on.
These figures take no account of the offsetting impact of extra investment.
If the capital stock is assumed to adjust, Mr Borjas reports, overall wages
are unaffected and the loss of wages for high-school drop-outs is cut to
below 5%.

Gianmarco Ottaviano, of the University of Bologna, and Giovanni Peri, of the
University of California, Davis, argue that Mr Borjas's findings should be
adjusted further. They think that, even within the same skill category,
immigrants and natives need not be perfect substitutes, pointing out that
the two groups tend to end up in different jobs. Mexicans are found in
gardening, housework and construction, while low-skilled natives dominate
other occupations, such as logging. Taking this into account, the authors
claim that between 1980 and 2000 immigration pushed down the wages of
American high-school drop-outs by at most 0.4%.

None of these studies is decisive, but taken together they suggest that
immigration, in the long run, has had only a small negative effect on the
pay of America's least skilled and even that is arguable. If Congress wants
to reduce wage inequality, building border walls is a bad way of going about
it.

* ³The Labor Demand Curve Is Downward Sloping: Reexamining the Impact of
Immigration on the Labor Market². George J. Borjas. Quarterly Journal of
Economics, 2003

³The Evolution of the Mexican-Born Workforce in the United States². George
J. Borjas and Lawrence Katz. NBER Working Paper 11281, April 2005

³Is the New Immigration Really So Bad?². David Card. NBER Working Paper
11547, August 2005

³Rethinking the Gains from Immigration: Theory and Evidence from the US².
Gianmarco I. P. Ottaviano, Univerity of Bologna and Giovanni Peri,
University of California, Davis, January 2006