37
largest (smallest) for those in sectors that use low-skilled labor more (less) intensively. Inflows of any
type of high-skilled labor (
0
ˆ
,
0
ˆ
2
1
>
> K
K
) will raise real wages of low-skilled workers while reducing real
wages of all high-skilled workers.
Unlike the basic “insensitivity” results from the HO model, however, these distributional effects
are compromised by the inclusion of non-traded goods in the model. If X
2
is a non-traded commodity
(e.g., medical care), any inflows of workers with skills specific to its production (K
2
) that generates an
increase in the production of X
2
can lead to a decline in P
2
– this can occur when consumption tastes
among individuals are such that the expansion in the output of X
2
is not matched by the increase in
aggregate consumer demand for X
2
(if, say, immigrants have tastes biased strongly in favor of traded
goods). If this is the case, as is clear from (B-6) above, high-skilled workers in the traded sector may
actually benefit, in real terms, from immigration of high-skilled workers who have training specific to the
non-traded sector. Perhaps even more telling, if production in the non-traded sector is highly intensive in
low-skilled labor, the same kind of analysis leads to the conclusion that inflows of low-skilled immigrants
may lead to a subsequent decline in P
2
and the real wage effects for native low-skilled workers then
become ambiguous. This is clear from (B-8): since we cannot be sure that nominal wages for low-skilled
workers fall more quickly than does P
2
, native workers may actually benefit in real terms if their
consumption tastes are biased strongly in favor of the non-traded good.
More ambiguity enters when we consider the large country case. Assume again that both
commodities are traded, but now allow that (world) prices may change in response to the total amount of
each good produced in the economy. If production of X
2
is relatively intensive in low-skilled labor,
inflows of low-skilled labor due to immigration will increase output of X
2
and can lead to a decline in P
2
(and an increase in P
1
). Now the effects of these immigration flows on the real wages of all high-skilled
and low-skilled workers become ambiguous, and depend in part upon consumption tastes.
Convention