Border Battles: The U.S. Immigration Debates

• 1
• 2
• 3
• 4

A simple expression of this possibility is depicted in Figure 5, which we label the “Dynamic Economy” model.  We begin with the assumption that wages in all local labour markets are in equilibrium at the intersection of the original demand and original supply curve, the point shown by W₀ and R₀.  We then let the labour demand curve in a selected city shift outward, perhaps reflecting a resource discovery.  As a result, wages in that city might be expected to rise to W_h and resident employment to expand to R_h.  However, since immigrants are mobile and attentive to economic rewards, they select the city with the dynamic economy over all others.  When they do so, they cause the supply of labour curve to shift out (and become more elastic) at wages above W₀.  Employment expands from R₀ to R₁+M.  Although the resulting wage rate, W₁, is below the counterfactual wage W_h, it may nevertheless remain for some time above the initial level W₀, and therefore above the wage in cities that did not experience the positive demand shock.  Resident employment also expands from R₀ to R₁.

Figure 5 provides a framework for interpreting Goldin’s finding of a negative wage impact.  When she reports, “wages were depressed in cities having an increase … [in their] foreign born” (emphasis added), she is referring to the difference between W_h and W₁, not to a wage below W₀.  This subtlety is overlooked in many summaries of Goldin’s contribution.  Reviews of Goldin reported by Hatton and Williamson (1998: 170), Dolmas and Huffman (2004: 1129), and Graham (2004: 60) suggest that immigrants depressed wages to a level below W₀.  This was not the case.  Resident wages were high in cities with large numbers of immigrants.

The dynamic economy model assumes a demand shock unrelated to immigration.  A second mechanism that would explain the failure of immigration to have a depressing impact on wages is an induced demand shift, that is, a shift in demand that responds to the arrival of immigrants.  More foreigners or indeed more population from any source by itself should not mean lower wages or increased unemployment because the additional people not only supply labour but add to the demand for output in a closed economy.  Even if we relax the textbook assumption of a closed economy and allow local labour markets to import goods from another region to meet the expanding demand, this induced demand story will still be valid.  This is because some goods and services must be produced where they are consumed – restaurant meals, construction, and educational and medical services.  In addition, an open economy will respond to an increase in its labour force by expanding production of tradable commodities in which it has a comparative advantage.  Thus an exogenous entry of immigrants that produces an outward shift in the labour supply curve will prompt a positive response from the labour demand side of the market.  It is also conceivable that, with an open economy, the impact of immigration may increase resident wages if the expansion of local industry pursuing a comparative advantage also allows those firms to exploit economies of scale (Romer 1996; Brezis and Krugman 1996) or if strong complementarities between immigrants and residents are at work (Ottaviano and Peri 2005).

The dynamic and open economy models imply that immigration does not harm resident workers in the common-sense way in which the term “harm” is generally understood.  In the cases we examined, the wage either increases or remains the same and the employment of residents is not reduced.  The empirical evidence from 100 years ago suggests that wages of residents increased and the local employment of residents increased in the presence of heavy immigration.  The estimates reported by Goldin and by Hatton and Williamson of a wage setback for resident workers are only valid if we interpret them to suggest that resident wages would have risen even faster without immigration.⁹  Such a counterfactual estimate is problematic for at least two reasons.  If one is looking either for an explanation of political opposition to immigration in the past, as was Goldin or Williamson in an earlier article (1982), or if one is defending a proposal to restrict immigration in the present, what is likely to matter most is whether real wages fall in the presence of immigration not whether the growth of real wages is  slowed (Williamson 1982: 256).  Second, as articulated by Goldin and by Hatton and Williamson, the counterfactuals that they offer assume that there is no impact of immigration on the forward momentum of the economy.  Yet, there is every reason to suppose that without immigration, wage growth, even in these high-wage cities, might have been slower or even halted.  There is a broad consensus that immigration accelerated the rate of economic growth.  The key mechanisms emphasized in the literature are:
 

For a review of the literature see Carter and Sutch (1999: 319-332).  

In a dynamic economy there is also a more direct connection between immigration and growth.  The fact that immigrants opportunistically select local markets with the highest wages suggests that the ebb and flow of immigrants can act as a “governor” for the economy.  By reducing the rate of wage growth, the local business boom can continue without being choked off by explosive wage growth (Carter and Sutch 1999: 338-339).  Thus the fact that the wage does not reach the hypothetical level of W_h can be considered as good for the economy and as good for resident workers.

• 1
• 2
• 3
• 4