« Growth Policy and Poverty Reduction | Main | Brad Setser: "A Little Too late" »

Tuesday, November 20, 2007

Are Immigrants Good for Cities?

Giovanni Peri reviews research on how immigration impacts cities. The research shows that "the less-educated [are] relatively unaffected by immigration while highly-educated and houseowners gain from it":

Immigration and cities, by Giovanni Peri, Vox EU: Immigration is one of the new century’s ‘hot button’ issues. Whether it is Romanians in Milan or Mexicans in Los Angeles, the concern and the debate has intensified. Political campaigns are fought and won on the issue in several nations and it has shifted party politics in many others. Most of the concern and political backlash focuses on the large and increasing presence of immigrants in cities – especially low-education immigrants.

This is easy to understand from the facts. In the US and Europe, immigration is disproportionately directed to cities, especially large metropolitan areas. In the recent decade immigrant arrivals – currently around 1.25 million people per year – accounted for 40% of the US population growth and for 50-75% of the growth of its largest metropolitan areas. For instance, 27% of people residing in London are foreign-born, as well as 28% of people in New York and 17% in Paris, vis-à-vis much lower national averages (respectively equal to 9%, 12.1% and 10% for the UK, the US and France).

Percentage of Immigrants in Top US cities
  Population in millions Percentage of foreign-born
Overall US 299.3 12.1
Top 17 metropolitan areas 105.1 26.9
New York 18.8 27
Los Angeles 12.9 35
Chicago 9.6 15
Dallas 6.0 17.4
Philadelphia 5.8 7.9
Houston 5.5 19.8
Miami 5.4 36
Washington, DC 5.3 21.3

Source: U.S. Bureau of Census, July, 2006.

What are the real effects of immigration on wages, rents and local prices faced by the natives? Does immigration drive out the native population? In theory it can work either way. In a static Walrasian world with homogeneous workers, an inflow of workers tends to depress wages, drive up rents and push out natives. However if variety of skills, complementarities in production and agglomeration economies are important, immigration could raise productivity of natives; immigration can drive urban growth which then makes the cities more economically attractive. The matter cannot be settled by logic. Facts are needed.

Recent research One recent strand of research uses data from US cities and states to explore issues such as the response of natives to immigrants and the impact of immigration on the local economy. Recent research by David Card[1] and others is identifying important regularities that point to a positive productivity effect of immigrants on natives overall, and to an increase of average housing value in high immigration cities. These average effects, however, are accompanied by an important distributional component. Highly educated natives enjoyed the largest benefits while the less educated did not gain (but did not lose much either) from immigration to their cities. In a series of recent papers coauthored with Gianmarco Ottaviano[2], I use US data to analyze the impact of immigration on wages, rents and local prices faced by native workers accounting also for the response of natives to immigrants in the form of relocation.

Crowding out natives? From a pure accounting perspective, immigrants were responsible for about 50% of the population growth of the top 100 US metropolitan areas during the 1990s. However, if the inflow of immigrants caused an outflow of native workers (towards areas with low immigration), it would simply change the composition of a city but would not produce net population growth. Analyzing the response of native population to immigrant population across cities, we find instead that large inflows of immigrants over the period 1970-2005 were not associated with any reduction of native population growth. In fact, in most cases, large immigration flows were associated to larger population and employment growth for natives as well. A part of this positive correlation is due to the fact that booming cities attracted natives and immigrants alike. However, we also isolated an immigrant-specific 'pull factor' in each city, in the form of enclaves of earlier immigrants that preferentially attracted co-nationals in period of large outmigration from the country of origin. Even these 'pull-driven' inflows of immigrants were not related at all to outflows of natives, which disproves the theory of crowding out.

Effects on wages and house prices A second result emerges from the cross-city analysis and it is similarly robust and significant. The average wage of native workers and the average housing price increased significantly more in cities with large immigration flows than in cities with low immigration flows over the period 1970-2005. While part of this effect is also explained by the 'booming city' theory, isolating the immigrant-specific pull-driven variation we still find that a 1% increase in the share of foreign-born increased the average native wages by around 0.3-0.4%, and the average house prices (and rents) by 1%. These positive average effects, however, are accompanied by distributional effects. Analyzing the impact by education group we find that native workers with no high school diploma experienced a small reduction in wages and small increase in their rents as a consequence of immigration, while those with college education experienced a significant wage and rent increase.

Click to enlarge

Our explanation of the positive wage effects relies on an important mechanism that seems to be operating in cities as well as in the US economy as a whole, and is based on the fact that the skill composition of immigrants is complementary to that of natives. Foreign-born individuals in the US are over-represented among workers with low skills (no degree) and among those with very high skills (graduate degrees particularly in science and technology). On the other hand, foreign-born are under-represented among workers with high-school and some college education. Most American workers, therefore (70% of which have high school or some college education), do not compete with immigrants for similar jobs but benefit from their complementary productive tasks. Moreover, even at similar levels of education native and immigrant workers tend to specialize in different occupations.

For instance, in related research[3] I found that among less educated workers, immigrants specialize in manual intensive tasks such as cooking, driving and building while natives specialize in language-intensive tasks such as dispatching, supervising and coordinating. The productivity of supervisors, clerks and accountants benefits from the productive services of construction workers, hand packers and janitors in their company. Similarly at high levels of education foreign-born specialize in analytical-mathematical tasks and natives in managerial-language intensive tasks. Again, lawyers’ productivity benefits from the competence of their computer and information technology assistants. Such skill differences translate into limited competition in the labour market and rather complementarities, inducing higher demand (and productivity) for native skills in economies where the supply of immigrants increases. As these complementarities work within as well as across education groups, there is a benefit for natives overall in producing in an economy with immigrants (the positive average wage effects). However the largest benefits are for those with intermediate and high education that do not compete at all for jobs with the large group of less educated immigrants.

On the other hand, this higher productivity of natives overall and the increased city population generates upward pressure on housing prices and rents. These effects, in the long run, induce more house building, so that their impact on house prices should not be large. This is true for housing of less educated workers who live in less desirable locations whose supply is very large. However, more educated individuals, who concentrate in highly desirable urban locations (e.g. Manhattan, Santa Monica or Downtown San Francisco), face a space constraint and experience increasing house values even in the long run.

Interestingly, our quantitative estimates imply an almost exact wash between the increase in average wages and that in average rents, at least on average, so that the real local wage, corrected for local prices is left essentially unchanged[4]. This can be interpreted as an effect of native’s mobility as they 'arbitrage away' real wage differences by moving across cities. The productive effect of immigrants ultimately accrues to house-owners who see the value of their houses increase.

Other effects The positive average wage effect combined with a somewhat adverse distributional effect and a positive house price effect (mostly for highly educated) leaves the less-educated relatively unaffected by immigration while highly-educated and houseowners gain from it. Where, therefore, does the vastly negative reaction to immigration come from? Two important channels, relative to local public good provision, are left out of our analysis. First, fiscal cost of immigrants at the local level may be relevant, especially if they use public goods (such as hospitals and schools) more than natives do and, because of their income, they contribute less in local taxes. Second, people seem to attribute positive value to living in ethnically homogeneous neighborhoods and to sending children to school with better-educated high-income families. Moreover, there may be important peer effects in education and learning. Opinion polls suggest that people may care significantly about these aspects of immigration.

Bottom line Our research shows that market-mediated economic effects of immigrations are mostly positive for natives. However, in order to affect policies it is also crucial to understand better the welfare-related effects and the peer effects that may explain most of the negative attitude towards immigrants.


1 David Card (2007) “How Immigration Affect U.S. Cities” CReAM Discussion paper #11/07, June 2007.
2 G.I.P. Ottaviano and G. Peri (2007) “The Effects of Immigration on U.S. wages and rents: a general equilibrium Approach” CEPR Discussion Paper # 6551; G.I.P. Ottaviano and G. Peri (2006) “The Economic Value of Cultural Diversity: Evidence from U.S. cities” Journal of Economic Geography, Vol. 6, Issue 1, Pages 9-44; and G.I.P. Ottaviano and G. Peri (2005) “Cities and Cultures” Journal of Urban Economics, Volume 58, Issue 2, Pages 304-307.
3 G. Peri and C. Sparber (2007) “Task Specialization, Comparative Advantages, and the Effects of Immigration on Wages” NBER Working Paper, # 13389, September 2007.
4 As reported above an increase in the share of immigrants by 1% increases rents by 1% and wages by 0.3-0.4%. As housing cost account for about 30% of total family expenditure the change in the local price index due to immigration is 0.3*1%=0.3% which is similar to the change in wages.

    Posted by on Tuesday, November 20, 2007 at 02:52 AM in Economics, Immigration, Unemployment | Permalink  TrackBack (0)  Comments (82)


    TrackBack URL for this entry:

    Listed below are links to weblogs that reference Are Immigrants Good for Cities?:


    Feed You can follow this conversation by subscribing to the comment feed for this post.