Journal of Ethnic and Migration Studies

Vol. 35, No. 7, August 2009, pp. 1129
1151

Immigrants and Neighbourhoods of

Concentrated Poverty: Assimilation or
Stagnation?
Paul A. Jargowsky

Immigrants in the United States often live, at least for a time, in neighbourhoods that
have high concentrations of fellow immigrants. Typically, these neighbourhoods also have
high poverty levels and are located near concentrations of the native-born poor.
Conventional wisdom is that living in extremely poor neighbourhoods leads to
‘concentration effects’ that exacerbate the problems of poverty and limit economic
opportunity. While immigrants are not immune to the problems of crime, gangs,
dilapidated housing and failing schools associated with high-poverty neighbourhoods, it
has been argued that immigrant neighbourhoods provide advantages as well. These
include the creation of parallel institutions, vernacular information networks and
familiar cultural practices. The analyses presented here provide some support for this
notion, by showing immigrants’ progress from higher- to lower-poverty neighbourhoods
over time. Yet Mexican immigrants do not transition nearly as rapidly, providing support
for the segmented assimilation hypothesis.

Keywords: Immigration; Concentration of Poverty; Barrios; Neighbourhood Effects;

Spatial Assimilation

Introduction
Racial and ethnic groups experience very different levels of neighbourhood socio-
economic disadvantage. A poor black person is over four times more likely to live in
an extremely poor neighbourhood than a poor person who is a non-Hispanic white.
The Hispanic poor are also much more likely than the white poor to live in high-
poverty neighbourhoods. Research on black ghettos would seem to demonstrate that
living in extremely poor neighbourhoods leads to poverty-enhancing ‘concentration

Paul Jargowsky is Professor of Public Policy at the University of Texas at Dallas. Correspondence to: Prof. P.A.
Jargowsky, School of Economic, Political and Policy Sciences, University of Texas at Dallas, 800 W. Campbell Rd,
Richardson, Texas 75080-3021, USA. E-mail: jargo@utdallas.edu

ISSN 1369-183X print/ISSN 1469-9451 online/09/071129-23 # 2009 Taylor & Francis

DOI: 10.1080/13691830903006150
1130 P.A. Jargowsky
effects’ that limit economic opportunity (Wilson 1987). Newly arriving immigrants,
particularly Mexicans, often spend some time living in very-high-poverty barrios,
or primarily Hispanic neighbourhoods. Does this have a negative effect on these
immigrants, or do these neighbourhoods actually serve a bridging function, so that
their benefits outweigh their costs? Prior research has provided conflicting theoretical
perspectives about the impact of concentrated immigrant neighbourhoods on their
residents. One school of thought is that beneficial aspects of immigrant clustering
outweigh the negative impacts of poverty concentration. Another line of argument is
that the concentration of poverty is just as harmful for low-income immigrants as it is
for the native-born poor, and serves to impede social and economic assimilation.
In the 1990s, the United States experienced a substantial reduction in the
concentration of poverty (Jargowsky 2003; Kingsley and Pettit 2003). The number
of high-poverty neighbourhoods*defined as census tracts with poverty rates of 40
per cent or more*declined by 26 per cent. The concentration of poverty, defined as
the percentage of a metropolitan area’s poor residents living in these economically
deprived neighbourhoods, declined for all racial and ethnic groups and in the vast
majority of metropolitan areas. The major exceptions were some metropolitan areas
in California such as Los Angeles-Long Beach, Fresno, Riverside-San Bernardino and
Bakersfield. These areas experienced a massive influx of low-income Mexican
immigrants, who expanded high-poverty zones and increased their populations.
These facts suggest that rapid immigration may explain why a number of
metropolitan areas in the South and West went against the prevailing national trend
of declining concentration of poverty despite the extremely strong economy of the
late 1990s. While the increasing concentration of poverty among Hispanics fuelled by
immigration is certainly a cause for concern, poverty researchers have traditionally
focused on high-poverty neighbourhoods in the Mid-West and North-East, asking
what effect residence in these neighbourhoods may have on the life-chances of their
residents, who are members of native-born minority groups. Often these residents
will have lived in high-poverty neighbourhoods for decades or even generations,
because they lack the means or the opportunity to leave (Quillian 1999, 2003). In fact,
many of the high-poverty neighbourhoods in the North-East and Mid-West were
formed by a process of selective out-migration of the non-poor, leaving the poor
behind (Gramlich et al. 1992; Jargowsky and Bane 1991).
In contrast, Latino barrios in the South and West are formed in part by movements
into these areas of low-income immigrants seeking employment opportunities
(Waldinger 1989). The social and economic dynamics of these neighbourhoods may
be quite different. They could represent more of a ‘gateway’ to residential and
economic mobility than the inner-city ghettos in other areas of the country, in which
case we should observe neighbourhood socio-economic assimilation over time
(Singer 2004; Yu 2004). A barrio may remain in place for decades, but the residents
observed at one point in time may move to better neighbourhoods, only to be
replaced by successive waves of immigrants (Lobo et al. 2002). The question is
whether individual immigrants use barrios as stepping-stones to more advantaged
 Journal of Ethnic and Migration Studies 1131

neighbourhood contexts or whether they become mired in high-poverty ones. The

evidence I present supports neighbourhood succession for most immigrant groups
but finds that progression from higher- to lower-poverty neighbourhoods is far
weaker among Mexican immigrants.

Are Barrios Different from Ghettos?

The study of high-poverty inner-city neighbourhoods in the United States has had a
heavy emphasis on African-American neighbourhoods, particularly in New York,
Chicago and other North-Eastern and Mid-Western cities. Given that African-
Americans have had much lower family income and higher poverty rates than whites,
segregated black neighbourhoods also tend to be poor, or at least poorer than
majority neighbourhoods. For this reason, the term ‘ghetto’ became synonymous
with neighbourhoods of poverty, dilapidated housing and social disorganisation.
These connotations strengthened during the 1960s as the result of racial riots, a rapid
rise in violent crime and the emergence of the suburbs as an alternative form of
residential life. Generations of scholars came to regard high-poverty neighbourhoods
as a social problem with consequences for child development and young adult
decision-making (Brooks-Gunn et al. 1997).
Given the low income of most new immigrants, they tend to settle in high-poverty
neighbourhoods where they have access to low-cost housing. As a result, they often
reside in close proximity to existing concentrations of the native-born poor, who are
disproportionately African-American (Portes and Zhou 1993). Segregated low-
income Hispanic neighbourhoods in the United States are traditionally called barrios.
Although the term translates simply as ‘neighbourhood’ or ‘village’ in most of the
Spanish-speaking world, within the US, the term implies a low-income neighbour-
hood in which the dominant language is Spanish. Barrios have ‘much in common’
with high-poverty African American neighbourhoods, so that the two are ‘sometimes
equated . . . as sites of neighbourhood disadvantage’ (Eschbach et al. 2004: 1807).
Yet the literature concerning immigrants living in ethnic neighbourhoods,
including barrios, discusses the opposite possibility. Forman (1971: 6) argued that
immigrant concentration was both voluntary and adaptive:

For so many different groups to have followed this practice for over a century
suggests that there are positive values to be derived from such concentration. For
the immigrants such voluntary segregation meant that they could reduce the rate of
transition from their native culture to the urban American way of life and thus
minimize ‘culture shock’. Immigrants established their own church with services in
their own languages, developed their own native-language newspapers, and
established their own clubs and mutual-aid societies. If the immigrant faced
difficulties, it was with the support of the centuries-old culture and traditions that
he and his neighbors had carried to the new American urban environment.

Immigrant neighbourhoods are often united by more than country of origin. A

chain of migration sometimes develops, in which immigrants from a particular
1132 P.A. Jargowsky
region, city or neighbourhood in the country of origin will settle in particular
neighbourhoods in the destination country (Massey et al. 1994). A ready-made set of
kinship and community ties is obviously a benefit to a newly arriving immigrant, and
can provide a significant incentive for cousins and friends to follow in the path of
earlier immigrants (Logan et al. 1994; Portes and Bach 1985).
Several studies have found evidence of the advantages conferred by immigrant
ethnic concentration, despite the poverty that naturally accompanies it. For example,
immigrant concentration has ‘sheltered newcomers from the ill-effects that nativist
activities may have had on their lives otherwise’ (Emeka 2006: 7). Within ethnic
neighbourhoods, new immigrants learned about industries and firms in which they
would be welcome from previous immigrants, who also gave them valuable referrals
and advice (Model 1988; Waldinger and Lichter 2003). For example, Hispanics in
barrios are more likely to rely on neighbours to learn about jobs than blacks in
ghettos (Elliott and Sims 2001). Eschbach et al. (2004), examining mortality rates
among older Mexicans, found that there is a ‘barrio advantage’ that counterbalances
the ‘negative health effects of neighbourhood disadvantage’. A study of immigrants in
Sweden, exogenously assigned by a state housing programme to different locations,
found a 13 per cent advantage in earnings for less-skilled immigrants living in ethnic
neighbourhoods (Edin et al. 2003). Given such considerations, Walks and Bourne
(2006: 273), examining poverty neighbourhoods in Canada with high proportions
of immigrants, have called for ‘a more nuanced interpretation of segregation,
ghettoization, and neighbourhood dynamics’.
In contrast, some authors have argued that the beneficial impact of ethnic
neighbourhoods in the early twentieth century may no longer apply. Portes and Zhou
(1993: 83
5) describe a process of ‘segmented assimilation’ that departs from earlier
immigrant processes in several respects. First, in contrast to earlier waves of European
immigrants, newer immigrants have a non-white racial identity (despite Census
Bureau definitions that classify Hispanic Origin as an ethnicity rather than a race).
Second, the concentration of immigrant households in the urban core ‘puts new
minorities in close contact with native-born minorities’. Co-location serves to cement
the identification of new immigrants as racially different from the majority, and
‘exposes second-generation children to the adversarial subculture developed by
marginalised native youths’. Third, changes in the structure of occupations have
restricted the opportunities for upward mobility by reducing the relative proportion
of mid-level jobs available to less-well-educated workers.
Given these conflicting theoretical perspectives, research is needed on the
concentration of poverty among immigrants. This study addresses four distinct
issues, using data from the 1990 and 2000 US Censuses. First, do recent immigrants
experience a concentration of poverty at levels comparable to the native-born poor?
Second, does immigration explain the increasing concentration of poverty in
metropolitan areas with large immigrant flows? Third, do immigrants pass through
high-poverty neighbourhoods and achieve neighbourhood economic assimilation as
implied by the gateway model, or do they persist in residing in barrios over the long
 Journal of Ethnic and Migration Studies 1133

term? Fourth, do all immigrant groups achieve similar levels of spatial assimilation or
does segmented assimilation also extend to neighbourhood economic assimilation?

Immigration and the Concentration of Poverty

The official US poverty line compares a family’s income to a standard amount meant
to represent the cost of basic necessities. It does not have any geographic component,
so the location of a poor family and the poverty of its neighbourhood do not affect
whether an individual family is defined as poor. To operationalise the concept of the
concentration of poverty, it is necessary firstly to define a neighbourhood unit and
then to define how poor a neighbourhood has to be to qualify as high-poverty.
Census tracts are the only practical choice to serve as proxies for neighbourhoods in a
nationwide study. These are small, relatively homogenous areas covering the whole
nation with an average population of around 4,000 persons (White 1987). Consistent
with prior research, a neighbourhood will be designated a high-poverty area if 40 per
cent or more of the residents are poor based on the official US poverty measure
(Jargowsky 1997). This poverty level identifies a relatively small group of extremely
disadvantaged neighbourhoods that often exhibit high levels of social problems.

The Concentration of Poverty among Immigrants

Table 1 provides information both on individual poverty and on residence in high-
poverty neighbourhoods for all US citizens in 2000, based on their nativity and
citizenship status. For comparison purposes, data for several racial and ethnic groups
are also shown. Foreign-born includes all persons born outside the United States,
regardless of country of origin. Foreign-born persons (regardless of income) were
more likely to live in high-poverty neighbourhoods than the native-born by 4.1 to 2.7
per cent, but not as likely as African-Americans (9.2 per cent) or Native Americans
(9.7 per cent). Foreign-born persons who had become naturalised citizens were about
half as likely as non-citizens to live in such areas.
As one might expect, the foreign-born are more likely to be poor based on the
federal poverty guidelines, as shown in the far-right-hand column of Table 1. Nearly
18 per cent of the foreign-born are poor based on family income, compared to 12 per
cent of the native-born. The concentration of poverty*the proportion of the poor
living in high-poverty neighbourhoods*is far higher for the foreign-born than for
non-Hispanic whites (10.2 compared to 4.1 per cent) but not as high as among
African-Americans (18.6 per cent) or Native Americans (19.4 per cent). Poor
immigrants had about the same chance of residing in high-poverty neighbourhoods
as native-born poor persons generally. These levels may seem low, since they imply
that 90 per cent do not live in high-poverty neighbourhoods. Recall, however, that the
40 per cent threshold level identifies a small set of extremely disadvantaged
neighbourhoods. The average neighbourhood poverty rate experienced by a poor
immigrant in 2000 was 22.3 per cent, well above the national average rate of 12.4, and
 1134
P.A. Jargowsky
Table 1. Residence in high-poverty neighbourhoods by nativity, citizenship, poverty status and race/ethnicity, 2000
All persons Incomes below poverty Incomes above poverty

In In In Group
high-poverty % in high-poverty % in high-poverty % in poverty
Total areas high-poverty Total areas high-poverty Total areas high-poverty rate %

United States 281,422 7,952 2.8 33,900 3,487 10.3 239,982 3,712 1.5 12.4

Nativity and citizenship

Foreign-born 31,108 1,284 4.1 5,498 561 10.2 25,268 690 2.7 17.9
Naturalised 12,543 332 2.6 1,322 118 8.9 11,106 204 1.8 10.6
citizen
Not a citizen 18,565 952 5.1 4,176 443 10.6 14,162 486 3.4 22.8
Native-born 250,314 6,668 2.7 28,402 2,926 10.3 214,715 3,022 1.4 11.7

Race/ethnicity
Non-Hispanic white 194,514 1,914 1.0 15,414 639 4.1 174,372 800 0.5 8.1
Black 34,362 3,173 9.2 8,146 1,512 18.6 24,568 1,498 6.1 24.9
Native American 2,448 239 9.7 608 118 19.4 1,760 114 6.5 25.7
Asian 10,172 275 2.7 1,322 129 9.8 9,023 120 1.3 12.8
Hispanic 35,238 2,301 6.5 7,798 1,074 13.8 26,653 1,162 4.4 22.6

Note: Population figures in thousands.

Source: 2000 US Census, Summary File 3. Some persons do not have a poverty status, so poornon-poorBtotal.
 Journal of Ethnic and Migration Studies 1135

this figure includes all immigrants regardless of length of residency or country of

origin.
The level of concentration of poverty among the foreign-born population varies
dramatically from one metropolitan area to another. Table 2 shows the 25
metropolitan areas with the largest foreign-born populations. Los Angeles and New
York dominate this skewed distribution, with over three million foreign-born persons
each. Both have foreign-born concentration of poverty rates of around 15 per cent;
however, the native-born concentration of the poor is also around 15 per cent in Los
Angeles, whereas it is double that figure in New York. This huge gap is driven by the
demographic composition of the foreign- and native-born populations in those two
areas, as discussed below.
In Los Angeles, the citizen/non-citizen distinction is substantially more important
than in New York. More than 16 per cent of the non-citizen foreign-born poor live in
high-poverty areas compared to 10 per cent of the foreign-born poor who have
become citizens there. In contrast, the figures in New York are 15 and 13 per cent.

Table 2. Concentration of poverty among the foreign-born by Metropolitan Area, 2000

% Concentration of the poor

Foreign-born Foreign-born

MSA/PMSA Number Poverty rate% Total Non-citizen Citizen Native-born

1 Los Angeles, CA 3,423,581 21.3 15.1 16.4 10.2 14.7

2 New York, NY 3,124,065 19.8 14.3 14.9 13.1 30.5
3 Non-Metropolitan 1,687,319 23.3 7.5 7.3 8.1 5.0
4 Chicago, IL 1,415,699 12.7 3.7 3.5 4.3 16.3
5 Miami, FL 1,136,638 18.5 7.2 7.0 7.6 21.9
6 Houston, TX 851,854 20.3 2.3 2.3 2.0 7.9
7 Orange County, CA 843,143 16.6 0.5 0.4 1.2 0.8
8 Wash., DC.MD/VA/WV 826,687 10.6 0.9 0.9 0.7 9.8
9 Riverside, CA 606,226 20.1 9.4 10.1 7.1 7.2
10 San Diego, CA 599,345 18.8 11.7 13.4 6.0 7.7
11 Dallas, TX 587,683 19.0 3.1 3.2 2.0 6.4
12 Oakland, CA 570,063 12.1 4.5 4.0 6.1 7.1
13 San Jose, CA 569,433 10.1 0.0 0.0 0.0 0.0
14 San Francisco, CA 551,549 10.7 0.4 0.5 0.2 2.6
15 Boston, MA/NH 495,951 14.7 4.8 5.2 3.7 4.0
16 Phoenix, AZ 453,023 24.7 12.8 13.5 7.9 9.6
17 Atlanta, GA 420,596 14.7 2.7 2.7 2.8 12.7
18 Fort Lauderdale, FL 408,757 14.9 1.7 2.0 1.0 5.5
19 Nassau, NY 390,317 8.9 0.0 0.0 0.0 0.0
20 Newark, NJ 383,167 11.4 3.2 3.1 3.4 15.2
21 Philadelphia, PA/NJ 353,260 14.6 12.7 14.4 9.8 20.3
22 Bergen, NJ 350,268 10.4 1.5 1.4 1.8 6.8
23 Detroit, MI 333,384 13.3 7.7 8.0 7.2 10.7
24 Seattle, WA 329,544 14.3 3.7 3.0 5.7 2.0
25 Las Vegas, NV/AZ 257,140 15.1 0.5 0.5 0.5 0.9
1136 P.A. Jargowsky
In the remaining metropolitan areas, the highest levels of the foreign-born
concentration of poverty were in Phoenix (12.8 per cent), Philadelphia (12.7 per
cent), and San Diego (11.7 per cent). Among the foreign-born, the concentration of
poverty rate for non-citizens was usually higher than or comparable to the rate for
naturalised citizens. In nine of the top 25 metropolitan areas*including Chicago, New
York and Newark, where blacks comprise a large segment of the poor population*the
concentration of poverty among the foreign-born poor was five or more percentage
points lower than for the native-born poor. In contrast, there were no metropolitan
areas where the foreign-born rate was five percentage points higher than the native rate.
Overall, the foreign-born poor are not more concentrated, and in many areas are
substantially less concentrated, than the native-born poor. However, many immi-
grants settle in areas physically contiguous to the native-born poor concentrations.
Figures 1
3 contrast the locations of high-poverty neighbourhoods and immigrant
concentrations for the Los Angeles, New York and Chicago metropolitan areas
respectively. In Los Angeles, many high-poverty neighbourhoods are majority
foreign-born, but about an equal number are majority native-born. More strikingly,
a vast number of majority immigrant neighbourhoods are not high-poverty areas,
but are located in close proximity to them. New York exhibits a different trend, in
which few of the high-poverty neighbourhoods contain a majority of immigrants,
and where the areas of immigrant settlement are for the most part some distance
from the high-poverty areas. In Chicago, almost none of the high-poverty areas are
majority immigrant, but virtually all the majority immigrant neighbourhoods are
contiguous to high-poverty zones. As discussed below, the immigrants to these three
metropolitan areas come from different regions of origin, partly explaining the
different settlement patterns.
In summary, the concentration of poverty was less pronounced among the foreign-
born as a group in 2000 than one might have expected, at least when immigrants are
not disaggregated by year of arrival and country of origin. Not clear from the analysis
thus far, however, is whether the low average level of concentration among the
foreign-born is characteristic of the group generally or reflects an assimilation
process. One caveat to this conclusion is that immigrants often settle in neighbour-
hoods that, while not high-poverty in themselves, are in close proximity to high-
poverty neighbourhoods settled by the native-born poor. A question to be taken up in
later analysis is whether cohorts of immigrants, defined by year of entry into the US,
remain in concentrated poverty neighbourhoods or upgrade to better ones over time.
First, however, the impact of immigration on the extent of concentrated poverty will
be examined.

The Impact of Immigration on Changes in the Concentration of Poverty

While a small percentage of immigrants settle in high-poverty neighbourhoods, their
numbers are still large enough to affect the overall extent of the concentration of
poverty. Nationally, the overall number of residents of high-poverty ghettos, barrios
 Journal of Ethnic and Migration Studies
Figure 1. Percentage foreign-born and neighbourhood poverty rates: Los Angeles Metropolitan Area, 2000

1137
1138 P.A. Jargowsky

Figure 2. Percentage foreign-born and neighbourhood poverty rates: New York

Metropolitan Area, 2000

and slums declined by 2.4 million persons (23.5 per cent) between 1990 and 2000,
falling from 10.4 million to 8.0 million. However, the decline would have been even
larger had not new immigrants taken the place of persons who resided in such
neighbourhoods in 1990. Nearly 629,000 residents of high-poverty neighbourhoods in
2000 were new immigrants who arrived in the United States during the 1990s. There
were 2.7 million fewer native-born persons living in high-poverty neighbourhoods in

Figure 3. Percentage foreign-born and neighbourhood poverty rates: Chicago Metro-

politan Area, 2000
 Journal of Ethnic and Migration Studies 1139

2000 than in 1990 and 400,000 fewer foreign-born persons who arrived prior to 1990.
Other things being equal, without new immigrants, the total decline in population of
these high-poverty neighbourhoods would have been 3.1 million (30 per cent).
While the majority of metropolitan areas experienced declines, there were a
number that experienced increases in the number of residents of high-poverty ghettos
and barrios. Figure 4 shows the demographic source of these increases for the
metropolitan areas with the largest rise in high-poverty neighbourhood population.
Los Angeles had the largest increase by far*nearly 300,000 persons*both native-
born poor residing in high-poverty neighbourhoods and newly arriving immigrants.
The number of immigrants from prior cohorts living in high-poverty areas also
increased, but constituted a much smaller share of the total increase.
Six of the top ten metropolitan areas in terms of the size of the increase in high-
poverty neighbourhood population are located in California. In each of these, the
increase is composed of a combination of native-born and new immigrants, with a
much smaller role or even a decrease in the number of immigrants from earlier
cohorts, i.e. before 1990. In contrast, in the Washington DC metropolitan area, new
immigrants do not drive the increase in the high-poverty population. In McAllen,
TX, a metropolitan area close to the US/Mexico border, almost all the increase is in
new immigrants, partially offset by a decline in prior immigrants*the pattern that is
the most consistent with the idea of Latino barrios as gateway communities.
While many of the metropolitan areas with large increases in high-poverty
neighbourhood populations were destinations for immigrants, not all metropolitan
areas with a high flow of immigrants experienced an increase in the concentration of
poverty. Figure 5 shows the components of the change in the population of high-
poverty areas for the metropolitan areas with the largest overall influx of immigrants
between 1990 and 2000. New York, for example, had more than 100,000 new
immigrants who lived in high-poverty areas in 2000, partially offsetting a decline of a

Los Angeles, CA
Fresno, CA
Riverside, CA
Wash., DC/MD/VA/WV
Bakersfield, CA
San Diego, CA
McAllen, TX
Providence, RI/MA
Chico, CA
Middlesex, NJ

0 100,000 200,000 300,000

1990s immigrants
Prior immigrants
Native-born

Figure 4. Source of increase in high-poverty neighbourhood population: Metropolitan

Areas with largest increases
1140 P.A. Jargowsky

New York, NY
Los Angeles, CA
Chicago, IL
Miami, FL
Houston, TX
Wash., DC/MD/VA/WV
Orange County, CA
Dallas, TX
San Jose, CA
Atlanta, GA

-200,000 -100,000 0 100,000 200,000 300,000

1990s immigrants
Prior immigrants
Native-born

Figure 5. Source of change in high-poverty neighbourhood population: Metropolitan

Areas with largest immigrant influx

similar number of native-born persons and prior immigrants. While Chicago, Miami,
Houston and Washington DC all had many new immigrants, few of them settled in
high-poverty neighbourhoods.
The prior discussion oversimplifies the analysis to a certain degree because the
trends in the high-poverty residence of foreign-born and native-born persons may
not be independent. For example, if a neighbourhood has a poverty rate of 30 per
cent among the native-born, and then a sufficient number of poor immigrants
entered the neighbourhood, the average poverty rate of the neighbourhood could
be driven above the 40 per cent threshold. In that case, all the residents of the
neighbourhood, regardless of immigration status, would be added to the count of the
high-poverty neighbourhood population.
In fact, the reality is more complex. Nationally, the census-tract-level correlation
between the native-born and foreign-born poverty rates is a surprisingly low 0.495.
Immigrants with incomes above the poverty level actually tend to reduce poverty in
neighbourhoods with high native-born poverty rates. Table 3 addresses this issue by
showing the count of high-poverty census tracts and the population residing in them
when neighbourhoods are classified by the overall poverty rate, the poverty rate for the
native-born only, and the poverty rate for the foreign-born only. As might be expected,
more than twice as many census tracts are above the 40 per cent threshold based on the
foreign-born poverty rate compared to the overall poverty rate (5,489 compared to
2,510). However, it is also true that there are more census tracts above the 40 per cent
threshold using the native-born poverty rate (2,663 compared to 2,510). When
neighbourhoods are classified by the native-born poverty rate, 8.5 million persons
reside in high-poverty areas compared to 8.0 million using the overall poverty rate.
Therefore, immigrant settlement patterns actually attenuate the concentration of
poverty. In the 2,663 census tracts where the native-born poverty rate is above 40 per
cent, more than half have foreign-born poverty rates below 40 per cent and in 280
 Journal of Ethnic and Migration Studies 1141
Table 3. Concentration of poverty under alternative poverty rates
High-poverty areas defined by poverty rate:

Overall Native-born Foreign-born

United States
Census tracts 2,510 2,663 5,489
Population
Non-Hispanic 7,952,260 8,456,013 20,342,696
White 1,913,547 1,900,511 11,460,925
Black 3,091,805 3,313,409 4,502,871
Hispanic 2,301,296 2,565,303 3,331,459

Los Angeles
Census tracts 137 189 116
Population
Non-Hispanic 559,574 793,846 463,443
White 35,472 46,261 37,678
Black 95,868 122,983 88,118
Hispanic 378,471 550,768 292,870

New York
Census tracts 253 322 138
Population
Non-Hispanic 944,865 1,246,530 482,823
White 89,093 116,946 47,302
Black 346,661 450,755 186,766
Hispanic 463,878 615,589 222,459

tracts the lower poverty of the foreign-born is sufficient to bring the overall poverty
level below 40 per cent. While 127 census tracts are also pushed above the 40 per cent
threshold by poverty among immigrants, the former effect is more pronounced. An
even more dramatic pattern is observed in Los Angeles and New York, where there are
substantially more census tracts that would be classified as high-poverty neighbour-
hoods by the native-born poverty rate than by either the overall or the foreign-born
poverty rate.
In summary, immigrants did contribute to the increasing population of high-
poverty neighbourhoods in metropolitan areas such as Los Angeles. Moreover, in
metropolitan areas like New York and Chicago, immigrants offset declines in the
population of high-poverty areas due to the out-migration of the native-born and
earlier waves of immigrants. However, the immigrant poor also settle in a less-
concentrated pattern than the native-born poor and actually reduce the measured
level of the concentration of poverty.

Neighbourhood Socio-Economic Assimilation

Immigrants and members of racial minority groups must overcome many hurdles to
fully assimilate into the social and economic mainstream of society. While the exact
1142 P.A. Jargowsky
meaning of assimilation and the process by which it occurs has been a controversial
topic, most agree that it is a multidimensional phenomenon that occurs over several
generations (Gordon 1964). Its spatial aspects were first highlighted by Robert Park
(1926), who argued that ‘residential mobility follows from the acculturation and
social mobility of individuals and that residential mobility is an intermediate step on
the way to structural assimilation’ (Alba and Nee 1997: 836). On the other hand, the
lack of spatial assimilation can be a bar to residential and social mobility and make
structural assimilation exceedingly difficult (Massey and Mullan 1984: 837). This
‘chicken and egg’ relationship is further complicated by several distinct definitions of
what constitutes spatial assimilation. For example, the term is sometimes used to
refer to residential integration with the majority group and at others to refer to
having access to desirable neighbourhoods and high-quality housing, even if those
neighbourhoods are still segregated (Alba and Nee 1997).
In this paper, I examine a single aspect of spatial assimilation, namely the ability of
immigrants to move from higher- to lower-poverty neighbourhoods. I do not take a
position here on whether such movements are the cause of or are caused by economic,
social or cultural assimilation. My concern is to document whether it is occurring, and
if so, how rapidly and for which immigrant groups. Whether cause or consequence,
the spatial assimilation of immigrants into less-poor neighbourhoods would be an
indicator of progress toward a broader conception of assimilation, and the lack of it
over decades or generations would indicate that the assimilation process had stalled.
The foreign-born are a diverse group, arriving in the US under different
circumstances and with highly varied human and social capital. Ideally, we would
fully disaggregate the foreign-born by region of origin, year of entry and poverty
status, among other variables. However, the census-tract data needed to study poverty
at the neighbourhood level impose some restrictions on this type of analysis. Summary
File 3 (SF3) of the US census is needed to identify high-poverty areas at the census-
tract level. SF3 comes in aggregated form as tables summarising variables for each
geographic unit, not as individual data. The Public Use Microdata Sample (PUMS)
does provide individual-level data and is often used for research on immigrants, but it
does not identify census tracts due to confidentiality concerns. So the PUMS cannot be
used to study the concentration of poverty among immigrants. Therefore, we are
limited to the tables as they are presented in the summary data, and it is not possible to
create new cross-tabulations or change the categories from those given.
Unfortunately, the 1990 census does not include a table comparable to the 2000
Census table (PCT 51) that gives poverty status by place of birth and citizenship
status. Moreover, the tables providing information on year of entry and country of
origin are not cross-tabulated with poverty status. Thus, in the analysis that follows, I
can only present figures on the proportion of all members of a particular group
residing in high-poverty zones, regardless of poverty status, not the proportion of the
poor, as I would prefer. Generally, the figure on residence in high-poverty areas is two
to four times higher for the poor members of a group than for the group overall.
 Journal of Ethnic and Migration Studies 1143

Race and Ethnicity

Table 4 shows the number and proportion of immigrants living in high-poverty areas
as well as the average neighbourhood poverty rate by nativity and citizenship status
for racial and ethnic subgroups. There are striking contrasts between immigrants by
race. For example, non-Hispanic white foreign-born persons almost never live in
high-poverty neighbourhoods, and the average poverty rate for this group is 10 per
cent, below the national average of 12.5 per cent. Among blacks, the foreign-born are
substantially less likely to live in high-poverty neighbourhoods (5.4 compared to 9.5
per cent). Nevertheless, they are nearly five times more likely to live concentrated
poverty areas than the white foreign-born, and on average they experience much
higher neighbourhood poverty levels (about 18 per cent). Hispanic foreign and
native-born, at 6.3 and 6.7 per cent respectively, are about equally likely to live in

Table 4. Neighbourhood poverty by race/ethnicity, nativity and citizenship, 2000

No. in % in Average
high-poverty high-poverty neighbourhood
Race/ethnicity Total areas areas poverty rate

All races/ethnicities 281,421,906 7,952,260 2.8 12.5

Foreign-born 31,107,889 1,283,993 4.1 15.6
Naturalised citizen 12,542,626 322,278 2.6 12.9
Not a citizen 18,565,263 951,715 5.1 17.4
Native-born 250,314,017 6,668,267 2.7 12.2

Non-Hispanic white 194,514,140 1,913,547 1.0 9.8

Foreign-born 6,840,532 80,967 1.2 10.0
Naturalised citizen 3,729,229 32,788 0.9 9.2
Not a citizen 3,111,303 48,179 1.5 11.0
Native-born 187,673,608 1,832,580 1.0 9.8

Black 34,361,740 3,173,016 9.2 20.9

Foreign-born 2,099,865 112,721 5.4 17.7
Naturalised citizen 934,367 41,671 4.5 16.8
Not a citizen 1,165,498 71,050 6.1 18.5
Native-born 32,261,875 3,060,295 9.5 21.1

Asian 10,171,820 274,567 2.7 11.9

Foreign-born 7,012,202 184,100 2.6 12.1
Naturalised citizen 3,502,021 57,569 1.6 10.8
Not a citizen 3,510,181 126,531 3.6 13.5
Native-born 3,159,618 90,467 2.9 11.4

Hispanic 35,238,481 2,301,296 6.5 19.0

Foreign-born 14,157,817 885,187 6.3 19.8
Naturalised citizen 3,939,732 192,733 4.9 17.5
Not a citizen 10,218,085 692,454 6.8 20.7
Native-born 21,080,664 1,416,109 6.7 18.5

Source: 2000 US Census, Summary File 3, Tables P88, PCT63a-PCT63i.

1144 P.A. Jargowsky
high-poverty neighbourhoods, and have an average neighbourhood poverty level of
around 20 per cent. Of course, all these figures would be much higher if we could
examine poor members of these groups separately.

Period of Arrival
The 1990 and 2000 Censuses both identify foreign-born persons by year of entry into
the United States, although the years are grouped differently, with more detail given
in 1990. Redstone and Massey (2004) have shown that census respondents often
interpret this question in different ways, especially in the context of multiple visits to
the US and returns to their native country. Nevertheless, they also show that errors
occur in both directions and analyses based on year of entry do not show a particular
bias in one direction or the other due to mis-reporting. The 1990 table gives more
detail before 1970, whereas the 2000 table has only one category for all immigrants
who entered before 1970. Nevertheless, three comparable cohorts of immigrants can
be constructed, allowing a weak form of longitudinal analysis: before 1970, 1970
79
and 1980
89.
Table 5 combines information from both sources and shows the number of
foreign-born persons by their year of entry into the United States and the proportion
residing in extremely poor neighbourhoods in 1990 and 2000. The table reveals a
wealth of information with just a few numbers. First, 5.3 per cent of all immigrants
lived in the highest-poverty neighbourhoods in 1990. Consistent with the national
trend in the 1990s, the proportion of immigrants living in concentrated poverty areas
declined to 4.1 per cent in 2000 (a 23 per cent decline in the proportion residing in
such neighbourhoods).
Looking down the 1990 column, we compare the residence in high-poverty
neighbourhoods by year of entry. We see a clear pattern of assimilation looking across
the cohorts. Fewer than 3 per cent of pre-1960 immigrants lived in high-poverty
zones in 1990, and more recent immigrants are consistently more likely to have lived
in high-poverty areas in 1990. The most recent group at the time of the 1990 census,
those who arrived from 1980 to 1989, lived in high-poverty areas 6.8 per cent of the
time. A similar pattern is observed as one reads down the year 2000 column.
Figure 6 shows the proportions living in concentrated poverty areas in 2000 for the
different arrival cohorts for 12 metropolitan areas. Regardless of year of entry,
immigrants are far more concentrated in some metropolitan areas*such as McAllen
TX and Fresno CA*than in other cities. These differences are probably driven by
differences in the country of origin of the immigrants in that area and by local
differences in housing markets. Despite very different levels across the cities, within
any given city, it is usually true that the longer a person has been in the United States,
the less likely he or she is to live in a high-poverty neighbourhood.
On the other hand, immigrants who arrived at different times might have differed
from each other prior to arrival. To rule out this possibility, we have to follow a given
cohort of immigrants (defined by year of entry) over time. Three comparable cohorts
 Journal of Ethnic and Migration Studies 1145
Table 5. Residence in high-poverty neighbourhoods by entry cohort, 1999
2000
1990 census 2000 census 1990
2000

Total % residing in Total % residing in

number high-poverty number high-poverty
Year of entry entering neighbourhood entering neighbourhood % change

Before 1950 1,842,688 2.9

1950
59 1,599,021 2.9
1960
69 2,792,565 3.4

Before 1970 6,234,274 3.1 4,778,099 2.1 31.90

1970
79 4,869,415 5.4 4,686,752 3.5 35.40
1980
89 8,663,627 6.8 8,464,762 4.6 32.90

1990
2000 13,178,276 4.8

Total 19,767,316 5.3 31,107,889 4.1 22.59

Source: 1990 and 2000 US censuses.

are shown in bold in the central panel reading across the rows of Table 5. For
example, the table shows that 6.2 million immigrants arrived before 1970. In 1990,
members of this cohort resided in high-poverty neighbourhoods 3.1 per cent of the

Bakersfield, CA El Paso, TX Fresno, CA Los Angeles, CA

McAllen, TX Minneapolis, MN/WI New York, NY Phoenix, AZ

Providence, RI/MA San Antonio, TX San Diego, CA Stockton, CA

0 .2 .4 .6 0 .2 .4 .6 0 .2 .4 .6 0 .2 .4 .6

All immigrants Entry before 1970

Entry 1970–79 Entry 1980–89
Entry 1990–2000

Figure 6. Concentration of poverty among immigrants, by year of entry: cities with more
than 100,000 immigrants, 2000. Graphs by Metropolitan Area (MSA/PMSA)
1146 P.A. Jargowsky
time, but this figure had declined by nearly a third by 2000. Note that the 2000 census
only registers 4.8 million persons as having arrived in the US before 1970. The major
decline in the size of this cohort probably reflects mortality; those who immigrated
before 1950 would have been 70 years old by 2000 if they were 20 years old on arrival.
Some may therefore have returned to their country of origin, perhaps to retire.
The next cohort are those who arrived in the 1970s. Of this group, 5.4 per cent
resided in high-poverty neighbourhoods in 1990, falling about a third to 3.5 per cent
by 2000. Of those immigrating in the 1980s, the high-poverty neighbourhood
percentage dropped by a third from 6.8 in 1990 to 4.6 in 2000. The size of the more
recent cohorts drops only slightly; presumably, mortality is a much smaller factor for
the more recent immigrants.
There is some evidence in Table 5 supporting the assimilation of immigrants who
initially live in high-poverty zones, but also some evidence against it. On the positive
side, we see that, at a given point in time, more recent cohorts are less likely to live in
neighbourhoods of concentrated poverty. We also see that a given cohort is less likely
to live in such a neighbourhood over time. Both considerations support a conclusion
that residence in high-poverty neighbourhoods is a transitory phenomenon and that
immigrants assimilate geographically to neighbourhoods of lower poverty levels
despite initial high-poverty neighbourhood residence.
On the other hand, the concentration of poverty was declining for all groups
between 1990 and 2000, regardless of immigration status, so we also need to take into
account period effects. We can do this, for example, by comparing along diagonals:
the newest immigrants in 1990 (those who arrived in the 1980s) compared to the
newest immigrants in 2000 (those who arrived in the 1990s). There is a comparable
reduction in the tendency to reside in high-poverty neighbourhoods, dropping from
6.8 per cent for the 1980s cohort in 1990 to 4.8 per cent for the 1990s cohort in 2000,
a 29 per cent reduction. However, comparing the 1970s cohort in 1990 to the 1980s
cohort in 2000 shows a 15 per cent reduction. The existence of such a strong period
effect somewhat calls into question the evidence of assimilation derived from
following individual cohorts from 1990 to 2000.

Region of Origin
The evidence above aggregated foreign-born persons from all nations of origin. Yet
immigrants from different continents and nations bring vastly different amounts of
personal and cultural capital with them. They have to go through different processes
and filters to gain entry, and are treated differently once they arrive (Portes and
Rumbaut 2001). Perlmann argues that national origin is crucial to understanding the
‘networks, niches, modes of incorporation, historical context, and the like’ that affect
the assimilation of different immigrants groups (2005: 8).
Figure 7 shows the average neighbourhood poverty rate in 2000 of immigrants
from seven broad regions of origin and three periods of arrival in the United States.
European immigrants lived in the least-poor neighbourhoods, and Mexicans in the
 Journal of Ethnic and Migration Studies 1147

25.0%

20.0%

Before 1980
15.0%
1980 – 89
10.0%
1990 – 2000
5.0%

0.0% ia

a
pe

o
an

a
.
m
ric

ic

ic
As
ro

be

.A
ex

er
Af
Eu

Am
ib

en
ar

C
C

S.
er
th
O
Figure 7. Average neighbourhood poverty rate in 2000, by region of origin and period of
immigration

poorest. All immigrant groups except Mexicans showed a pattern that could be
considered as spatial assimilation: the immigrants from earlier periods live in
progressively less-poor neighbourhoods. The trend is even more dramatic for the
proportion who lived in concentrated poverty neighbourhoods, as shown in Figure 8.
In fact, a higher proportion of recent Caribbean immigrants lived in the poorest
neighbourhoods, but the rate is much lower for Caribbean immigrants who have
been around longer. In contrast, Mexican immigrants had about the same probability
of high-poverty neighbourhood residence, and the same average neighbourhood
poverty rate, regardless of period of entry.

9.0%
8.0%
7.0%
6.0% Before 1980
5.0%
1980–89
4.0%
3.0% 1990–2000
2.0%
1.0%
0.0%
pe

ia

an

a
.
m
ric

ic

ic
As
ro

be

.A
ex

er
Af
Eu

Am
ib

en
ar

C
C

S.
er
th
O

Figure 8. Probability of living in a high-poverty neighbourhood in 2000, by region of

origin and year of entry
1148 P.A. Jargowsky
Conclusion
The large volume of immigration in recent decades, particularly that from Mexico,
has aroused strong passions in the American electorate. Opponents of immigration
contend that recent immigration from Mexico and Central America is so massive that
it will breed a Latino underclass and ‘defeat the American tradition of assimilation’
(MacDonald 2004). On the other hand, immigration can also provide support for
neighbourhoods that otherwise would have been depleted by out-migration and
abandonment (Muller 1993; Winnick 1990).
The reality is complex. Immigrants in general experience more concentration of
poverty than non-Hispanic whites, but much less than do minority groups who
constitute the bulk of the native-born poor in most urban areas. The impact of
immigration on inner-city poor neighbourhoods is more ambiguous than many
people assume and varies dramatically from one metropolitan area to another. The
immigrant poor are less concentrated than the native-born poor, and when the two
settlement patterns are overlaid, as they often are, the measured level of the
concentration of poverty is actually reduced. In other words, while immigration
undoubtedly raises poverty rates in some areas, it also tends to reduce poverty rates in
many of the poorest inner-city communities.
A key issue addressed here is the extent to which immigrant populations
experience geographic assimilation, i.e. do they migrate from higher- to lower-
poverty neighbourhoods over time? The Chicago School argued that population
groups would seek to convert economic and social gains into spatial amenities by
moving to neighbourhoods at higher levels in the urban hierarchy (Park 1926). More
recently, there has been a focus on the role of spatial patterns in facilitating or
hampering social mobility by affecting the ability of a group to access public services,
economic opportunities and information networks (Jargowsky 1997; Massey 1990;
Massey and Denton 1993; Wilson 1987). Either way, the degree of spatial integration
of immigrant groups is a crucial indicator of the extent to which the American
experience of assimilation continues among more recent waves of immigrants. The
data presented here suggest that, while most immigrants experience assimilation in
terms of neighbourhood poverty rates, Mexican immigrants seem to remain in
impoverished neighbourhood contexts for decades. Thus, the experience of Mexican
immigrants, at least those who arrive poor and initially settle in high-poverty barrios,
is consistent with the segmented assimilation hypothesis.
Prior theoretical and empirical research has suggested that the concentration of
immigrants can serve a useful purpose, helping to ease their transition into a new
society, especially for those who do not speak the language of the dominant group.
While immigrants are not immune to the problems of crime, gangs, drug and alcohol
abuse, dilapidated housing and failing schools that plague high-poverty areas, it has
been argued that immigrant neighbourhoods provide advantages as well, at least for
earlier generations of immigrants. These include the creation of parallel institutions,
vernacular information networks and familiar cultural practices that could ease the
 Journal of Ethnic and Migration Studies 1149

transition to the new society. The analyses presented here provide some support for
this notion overall, by showing that many immigrant groups move from higher- to
lower-poverty neighbourhoods over time.
At the same time, Mexican immigrants do not appear to experience neighbour-
hood socio-economic assimilation, at least not at the same rate as immigrants from
other countries and regions of origin. The strong suggestion is that the concentration
of poverty among certain immigrant groups, particularly those who arrive with low
levels of human capital, can have the same types of negative consequence as have long
been noted in the neighbourhoods of the native-born poor. Much more research is
needed on the neighbourhood effects in immigrant communities, particularly on the
mechanisms that seem to protect some immigrant groups while leaving others
vulnerable. Another question with important policy implications is the extent to
which the children of immigrants are affected when they reside with their parents in
high-poverty immigrant enclaves. A full appreciation of the consequences of the
concentration of poverty among immigrants and the implications of those effects for
economic and social assimilation will require an understanding of the effects of poor
neighbourhoods on the second and subsequent generations.

Acknowledgements
An earlier version of this paper was presented at ‘Local Contexts and the Prospects for
the Second Generation’, in Seattle, Washington, 20 October 2006. The West Coast
Poverty Center supported and encouraged this research. Helpful comments and
suggestions have been provided by Gunnar Almgren, Kurt Beron, Shauna Carlisle,
Marie Chevrier, Kyle Crowder, Mark Ellis, Magnus Lofstrom, Robert Plotnick,
Alejandro Portes, Hilary Silver and a JEMS referee.

References
Alba, R. and Nee, V. (1997) ‘Rethinking assimilation theory for a new era of immigration’,
International Migration Review, 31(4): 826
74.
Brooks-Gunn, J., Greg, J.D. and Aber, J. (eds) (1997) Neighbourhood Poverty. Vol. 1, Context and
Consequences for Children. New York: Russell Sage Foundation.
Edin, P., Fredriksson, P. and Aslund, O. (2003) ‘Ethnic enclaves and the economic success of
immigrants: evidence from a natural experiment’, Quarterly Journal of Economics, 118(1):
329
57.
Elliott, J. and Sims, M. (2001) ‘Ghettos and barrios: the impact of neighbourhood poverty and race
on job matching among Blacks and Latinos’, Social Problems, 48(3): 341
61.
Emeka, A. (2006) ‘Context of reception and the incidence of poverty among children of recent
immigrants’. Paper presented at the West Coast Poverty Center, University of Washington, 20
October.
Eschbach, K., Ostir, G., Patel, K., Markides, K. and Goodwin, S. (2004) ‘Neighbourhood context
and mortality among older Mexican Americans: is there a barrio advantage?’, American
Journal of Public Health, 94(10): 1807
12.
Forman, R. (1971) Black Ghettos, White Ghettos, and Slums. Englewood Cliffs, NJ: Prentice Hall.
Gordon, M. (1964) Assimilation in American Life. New York: Oxford University Press.
1150 P.A. Jargowsky
Gramlich, E., Laren, D. and Sealand, N. (1992) ‘Moving into and out of poor urban areas’, Journal of
Policy Analysis and Management, 11(2): 273
87.
Jargowsky, P. (1997) Poverty and Place: Ghettos, Barrios, and the American City. New York: Russell
Sage.
Jargowsky, P. (2003) Stunning Progress, Hidden Problems: The Dramatic Decline of Concentrated
Poverty in the 1990s. Washington, DC: The Brookings Institution.
Jargowsky, P. and Bane, M. (1991) ‘Ghetto poverty in the United States: 1970 to 1980’, in Jencks, C.
and Peterson, P. (eds) The Urban Underclass. Washington, DC: The Brookings Institution,
235
73.
Kingsley, G. and Pettit, K. (2003) Concentrated Poverty: A Change in Course. Washington, DC: The
Urban Institute.
Lobo, A., Flores, J. and Salvo, J. (2002) ‘The impact of Hispanic growth on the racial/ethnic
composition of New York City neighbourhoods’, Urban Affairs Review, 37(5): 703
27.
Logan, J., Alba, R. and McNulty, T. (1994) ‘Ethnic economies in metropolitan regions: Miami and
beyond’, Social Forces, 72(3): 691
724.
MacDonald, H. (2004) ‘A new Latino underclass’, The Dallas Morning News, 25 July. Online at
http://www.manhattan-institute.org/html/_dmn-new_latino_underclass.htm.
Massey, D. (1990) ‘American apartheid: segregation and the making of the underclass’, American
Journal of Sociology, 96(2): 329
57.
Massey, D. and Denton, N. (1993) American Apartheid: Segregation and the Making of the
Underclass. Cambridge, MA: Harvard University Press.
Massey, D. and Mullan, B. (1984) ‘Processes of Hispanic and black spatial assimilation’, American
Journal of Sociology, 89(4): 836
73.
Massey, D., Goldring, L. and Durand, J. (1994) ‘Continuities in transnational migration: an analysis
of nineteen Mexican communities’, American Journal of Sociology, 102(6): 939
99.
Model, S. (1988) ‘Mode of job entry and the ethnic-composition of firms: early 18th-century
migrants to New York City’, Sociological Forum, 3(1): 110
27.
Muller, T. (1993) Immigrants and the American City. New York: New York University Press.
Perlmann, J. (2005) Italians Then, Mexicans Now: Immigrant Origins and Second Generation
Progress: 1890
2000. New York: Russell Sage Foundation.
Park, R.E. (1926) ‘The urban community as a spatial pattern and a moral order’, in Burgess, E.W.
(ed.) The Urban Community. Chicago: University of Chicago Press, 3
18.
Portes, A. and Bach, R. (1985) Latin Journey: Cuban and Mexican Immigrants in the United States.
New York: Cambridge University Press.
Portes, A. and Rumbaut, R. (2001) Legacies: The Story of the Immigrant Second Generation. Berkeley:
University of California Press.
Portes, A. and Zhou, M. (1993) ‘The new second generation: segmented assimilation and its
variants’, Annals of the American Academy of Political and Social Science, 530: 74
96.
Quillian, L. (1999) ‘Migration patterns and the growth of high-poverty neighbourhoods, 1970

1990’, American Journal of Sociology, 105(1): 1
37.
Quillian, L. (2003) ‘How long are exposures to poor neighbourhoods? The long-term dynamics of
entry and exit from poor neighbourhoods’, Population Research and Policy Review, 22(3):
221
49.
Redstone, I. and Massey, D. (2004) ‘Coming to stay: an analysis of the US census question on
immigrants’ year of arrival’, Demography, 41(4): 721
38.
Singer, A. (2004) The Rise of New Immigrant Gateways. Washington, DC: The Brookings Institution.
Waldinger, R. (1989) ‘Immigration and urban change’, Annual Review of Sociology, 15: 211
32.
Waldinger, R. and Lichter, M. (2003) How the Other Half Works: Immigration and the Social
Organization of Labor. Berkeley, CA: University of California Press.
Walks, R. and Bourne, L. (2006) ‘Ghettos in Canada’s cities? Racial segregation, ethnic enclaves and
poverty concentration in Canadian urban areas’, The Canadian Geographer, 50(3): 273
97.
 Journal of Ethnic and Migration Studies 1151

White, M. (1987) American Neighborhoods and Residential Differentiation. New York: Russell Sage
Foundation for the National Committee for Research on the 1980 Census, c1987.
Wilson, W. (1987) The Truly Disadvantaged: The Inner-City, the Underclass and Public Policy.
Chicago: University of Chicago Press.
Winnick, L. (1990) New People in Old Neighbourhoods: The Role of Immigrants in Rejuvenating New
York’s Communities. New York: Russell Sage Foundation.
Yu, Z. (2004) ‘Immigration and sprawl: residential location choice in three gateway metropolitan
areas of the United States’. Paper prepared for the 2004 ACSP Annual Conference in Portland,
Oregon, 21
24 October.