How Might Socio-Economic Aspects Impact the Settlement Patterns of Haitians and Jamaicans in U.S. Cities? Focus on New York and Miami Metropolitan Areas

Abstract

The topic of the mosaic patterns of the settlement of ethnic minorities in US cities has been studied in depth. Many works proceed from the optics of studying inequality, which, in the American context, is often explained by overlapping patterns of ethnoracial discrimination. In this regard, groups at the intersection of inequality patterns are of great interest—for example, non-white immigrants and their descendants. This article analyzes the settlement of two such groups—Haitians and Jamaicans—at the state, county, and census tract levels in the two main population centers for the following two groups: New York and Miami metropolitan areas. The authors pose the question of whether Haitians and Jamaicans fundamentally differ from African Americans in terms of settlement patterns and socio-economic status. It is established that these two population groups are highly concentrated on a nationwide scale (in the states of Florida and New York) and in some counties (comprising metropolitan areas of New York and Miami). The authors carried out a spatial correlation (LISA) analysis to identify areas of concentration of Haitians and Jamaicans in the two above-mentioned metropolitan areas. Three distinct Haitian areas, three mixed Haitian/Jamaican, and one distinct Jamaican area are found in New York. In the Miami metropolitan area, three mixed Haitian/Jamaican areas are identified, as well as one Haitian area. Calculation of the Darden–Kamel Composite Socio-Economic Index was used to assess the socio-economic status of these areas and to compare it with that in the areas of concentration of African Americans. The analysis revealed that the areas of concentration of Haitians and Jamaicans are relatively disadvantaged in their socio-economic status, but to a lesser extent than those of African Americans. The study also established the following pattern: in the New York metropolitan area, the higher socio-economic status is a feature of the mixed Haitian/Jamaican areas of concentration in the central city, while in Miami, the higher socio-economic status is observed in the Jamaican areas in the suburbs. Beyond these empirical findings, this article offers a contextual perspective on inequality among non-white migrant groups as they settle near established African American communities. The novelty of our approach to the research problem relates to the idea that communities of African Americans, Haitians, and Jamaicans are quite variable in time and space—in particular, due to the non-simultaneous settlement of these groups in the metropolitan areas under consideration. Our most important conclusion is the discovery of the relationship between the stage of the beginning of the integration of the groups considered into urban communities and their socio-economic status at that moment. The later the group began integration, the fewer discriminatory practices it had to face, which led to a higher socio-economic status.

1. Introduction

1.1. Research Relevance

Population in major cities in the United States is characterized by a great diversity of ethnic and racial composition. Despite the great potential for cultural exchange in the context of such diversity the problem of ethnic and racial inequality and the segregation of urban space has been quite acute throughout the history of the United States. Even after significant progress has been made toward achieving social justice, several groups remain in worse socio-economic conditions relative to the average. To understand the nature of inequality, it is important to study its spatial aspect, particularly the relationship between the concentration of social groups in certain residential areas that have an unfairly deprived social status.

When studying individual ethnic groups in the U.S. population, one should take into account ethnoracial context. Therefore, before proceeding to the statement of the problem, which a geographical study may solve, we should clarify ethnic and racial categories surrounding Haitians and Jamaicans in the United States. The division of the American society into racial groups implies the existence of African Americans. The fact of belonging to this group, as well as to any other racial group, is determined during census by means of a survey. A respondent answers voluntarily, therefore attributing themselves to a certain racial group, which is considered racial self-identification. Studies of the ethnic and racial composition of the U.S. population include another form of self-identification—the language form—into the number of racial groups. This is allowed exclusively for the Spanish-speaking population and corresponds to the racial group of Hispanics or Latin Americans [1].

This puts some groups of migrants in a dubious position, as they represent the Latin American region, but do not identify themselves as a part of the Spanish-speaking group. Haitians and Jamaicans are among the largest groups in this category. The situation is further complicated by the fact that despite racial groups being self-identified, attributing oneself to African Americans may or may not coincide with a societal perception of one’s race [2]. The perception of race and some specific types of it, most notably colorism, make a large contribution to the development of discriminatory politics, policies, and practices [3,4,5,6,7].

Representatives of Haitian and Jamaican communities in the U.S. may attribute themselves to African Americans or Latin Americans, even more so given that the latter category splits into White Hispanics and Black Hispanics in the detailed version of the census data. Therefore, it is important to understand which context is to be used when analyzing the following two groups: African American or Hispanic. While the countries of Haiti and Jamaica geographically belong to Latin America, migrants from these countries have an unclear status in the context of U.S. studies. On the one hand, they are most frequently included in the African American racial group, which is frequently based on colorist stereotypes. On the other hand, many studies emphasize the socio-economic and cultural differences between Black/African Caribbean or West Indian and African migrants and the U.S.-rooted African Americans. African Caribbeans are a part of African American society, as many people who identify themselves as African Americans have Caribbean roots [8]. New migrants, for example those arriving from Jamaica, are frequently included in the process of the assimilation and transformation of their identity into African American one [9]. Historically, Afro-Caribbeans faced the same manifestations of racial discrimination, including spatial segregation, as African Americans of non-Caribbean descent [10].

At the same time, there is rich evidence of African Americans perceiving Jamaicans and, especially, Haitians as migrants, that is, a group drastically different from them [11,12]. Jamaicans, Haitians, as well as other migrant groups in the U.S. try to settle in a more compact way, primarily in the conducive environment of unprivileged districts of the country’s major cities [13]. The settlement of Haitians and Jamaicans is often viewed by the researchers in the context of migration and diaspora studies [14,15,16,17] or together with that of migrants from the West Indies [11]. Many authors have concluded that Haitians and Jamaicans are heavily concentrated in the central cities of several major U.S. metropolitan areas, and their level of socio-economic well-being is substantially lower than the average. The settlement and socio-economic status of African Americans are in turn considered in tight connection with segregation issues. This discourse was established at the dawn of quantitative sociology, the first example being the pioneering study by Du Bois [18], and was continued by the proponents of radical geography [19,20,21].

1.2. Research Problem

The most significant contributions to the development of methodologies for studying the settlement and socio-economic status of African Americans were made by Massey, Denton [22], Wilson [23], and Johnston et al. [24], all these studies being a part of a multidisciplinary discourse formed by quantitative sociology, urban demography, and urban geography and treating segregation and socio-economic inequality as phenomena that could be quantified through indices. They also agreed that ghettos and other forms of poor neighborhoods have a dual, racial-class genesis. This means that such neighborhoods territorialize the intersection of patterns of inequality, rather than being a direct consequence of any singular factor.

The studies outlined in the previous paragraph contributed to a long-standing tradition of analyzing ethnoracial inequality in U.S. cities with an emphasis on spatial patterns. Conceptual approaches of this kind include assimilation theory, developed by Park and Burgess to study migrants in Chicago in the first half of the 20th century. They identified a pattern in which migrants, as they assimilated into American (at that time, predominantly Anglo-American and German American) society, improved their socio-economic status, simultaneously moving from poor areas of the central city to prosperous suburbs [25,26].

This model ceased to work after the two Great African American Migrations (1910s–1970s), as African Americans were subjected to unprecedented pressure from segregation institutions and other discriminatory mechanisms (in particular, in the housing market) [27]. Considering the violation of the pattern identified by Park and Burgess, Lee’s push–pull theory was adapted to analyze the settlement of non-white groups [28,29]. The central cities of metropolitan areas became increasingly marginalized and stigmatized during the 1950s–1980s due to the outflow of the wealthy white population, which could not be followed by the mass suburbanization of African Americans due to discrimination and poverty [30]. According to the push–pull theory, this was caused by an excess of a certain threshold value regarding two key factors, which became the lever for the mass relocation of the white middle class from central cities of metropolitan areas (white flight). The first factor is the increased attractiveness, accessibility, and affordability of quiet, traditional family-friendly suburbs, and the second one is the persistent racial prejudice that underpinned most of the discriminatory practices of the 20th century [31,32,33].

In most cases, they resorted to using different indices—dissimilation, concentration, centralization, and many other [34]. Even though the spatial inequality patterns of African Americans are much more studied than those of Haitians and Jamaicans, many basic singularities of their settlement and socio-economic status are like one another. All the three groups—African Americans, Haitians, and Jamaicans—are highly urbanized, concentrated in space, and relatively poor. Racial discrimination was historically a major factor that made an impact on the formation of these traits. As part of racist policies (such as redlining), the ethnic identity of individual groups was not accounted for in any way, so migrants from Haiti and Jamaica found themselves in the same conditions as African Americans who were born in the U.S. This is another factor allowing us to study them in one and the same context.

It should also be noted that, despite the importance of the American context and the specificity of the two groups under consideration, we believe that the optics proposed in this article can be applied to large cities in other countries as well. Thus, we can assume that similar patterns of socio-economic inequality between ethnic and other types of segregated communities are observed, for example, in Tokyo [35], Taipei [36], Paris [37], and the largest cities of South Africa [38,39].

The experience of studying inequality in cities in countries outside the United States is very rich and shows that, despite completely different principles of organizing patterns of inequality, it often comes in similar forms. At the same time, there are very interesting exceptions to the trends that have developed in the 20th and 21st centuries, for example, Kowloon Walled City in Hong Kong, which was dramatically overpopulated because of the lack of clear regulation and planning. Its unregulated, self-organized structure led to natural divisions based on socio-economic status, functional balance, and spatial accessibility. Although formally under Chinese jurisdiction, it served as a lawless enclave, shunned by most Hong Kong residents as a hub of crime and vice. Its residents faced marginalization, with limited access to formal economic opportunities, education, and public services, in contrast to the modernizing urban environment surrounding it. Chaotic, self-contained design and a lack of government oversight further isolated it physically and symbolically, creating a sharp divide between its infrastructure and Hong Kong’s regulated development [40,41].

There is a knowledge gap today, because research on this issue has tended to focus either on African Americans as a racial group (which is debatable) or on one or another group of non-white migrants without regard to non-white groups that have lived in the United States for a longer period. Their urban concentration patterns are often adjacent and are the product of similar discriminatory practices, so we find it useful to place them in the same research context.

Thus, the main research question of this study is what the settlement patterns and socio-economic statuses of Haitians and Jamaicans have in common with African Americans and what differences they have. As these two ethnic groups were not directly analyzed using the means of quantitative socio-economic geography and quantitative sociology, the relevance of this study is also connected with filling the gap in this area of American urban studies. Thus, the main goal of this paper is to analyze the settlement patterns of Haitians and Jamaicans in U.S. cities in the context of inequality and segregation studies as well as quantitative African American studies. By comparing these two ethnic groups to a much larger group of African Americans (which is, of course, wider than an ethnic group, being both a racial and a socio-economic group because of the tragic history of racial discrimination and artificial inequality), we can find out whether Haitians and Jamaicans can be analyzed as an integral part of the African American group or whether they are more of an exception.

2. Materials and Methods

The data were collected from the American Community Survey for the years 2014 through 2018, as conducted by the U.S. Census Bureau [42]. It contains tables with data on the respondents’ ancestry, including the options “Haitian” and “Jamaican”. Notably, the perception of ancestry does not set a framework for migrant generations. A Jamaican person may perceive themselves as either having been born in Jamaica and not having U.S. citizenship or as a third-generation descendant of migrants. Therefore, the categories of “Haitian” and “Jamaican” in this context may be contrasted with the “African American” category as an ethnic rather than a racial self-identification option.

The socio-economic data are also contained in the American Community Survey. Below, as part of the methodology description, we will name the parameters collected from that source. The collection of data was conducted not directly on the Census Bureau website, but from the project website of the Minnesota Population Center at the University of Minnesota, NHGIS [43], as the latter accumulates the official Census Bureau’s data in a more convenient way. All the data were imported into the QGIS software (version 3.34) in a comma-separated values format and were then geocoded.

Apart from the analysis of the size of the Haitian and Jamaican population in states and counties, an analysis was carried out on the level of census tracts (CTs) in metropolitan areas with large Haitian and Jamaican communities. This was conducted in two stages:

• Identifying areas of concentration of the Haitian and Jamaican population (as well as African Americans—for the purposes of comparison).
• Assessing their socio-economic status (also as compared to African Americans).

2.1. Identifying Areas of Concentration

At the first stage, we applied the spatial correlation analysis algorithm. It was proposed by Luc Anselin based on the analysis of local spatial autocorrelation [44]. Spatial autocorrelation is a quantitative relationship between the objects in question and the congruence of the values of certain parameters in them [45]. It may be positive (when neighboring objects have similar values) or negative (when neighboring objects have different values).

To assess the spatial autocorrelation, we need to calculate Moran’s I, which may be Global or Local. Global Moran’s I assesses spatial autocorrelation for the totality of the objects, while Local Moran’s I assesses each object individually. The Global I was proposed by Patrick A. P. Moran in 1950 for biometric studies [46], while the Local I index was devised by Anselin in 1995 specifically for the analysis of ethnic group concentrations [44].

If the analysis is based on two parameters instead of one, such a procedure may be called the spatial correlation analysis (in Anselin’s terms, it is Bivariate Local Moran’s I). To conduct the spatial correlation analysis, we used the Nearest Neighbor Analysis algorithm. It assumes the following formula:

$${\mathrm{I}}_{\mathrm{i}}={\displaystyle \frac{{\mathrm{x}}_{\mathrm{i}}-\overline{\mathrm{x}}}{\frac{\sum _{\mathrm{j}=1,\mathrm{i}\ne \mathrm{j}}^{\mathrm{n}}{\mathrm{w}}_{\mathrm{i}\mathrm{j}}}{\mathrm{n}-1}-{\overline{\mathrm{x}}}^2}}\times \sum _{\mathrm{j}=1,\mathrm{i}\ne \mathrm{j}}^{\mathrm{n}}{\mathrm{w}}_{\mathrm{i}\mathrm{j}}\times \left({\mathrm{y}}_{\mathrm{j}}-\overline{\mathrm{y}}\right),$$

where n is the number of census tracts, w_ij is the weighting matrix element, W is the sum of all w_ij, x_i is the minority population in i, y_j is the share of the minority population in j, and $\overline{y}$ is the mean share of the minority population in all census tracts.

The spatial correlation analysis has an advantage over the spatial autocorrelation analysis in cases where census tract-level statistical data are used. The use of two variables allows distortions connected with the engineering nature of the statistical–territorial division to be avoided. For example, we often encounter cases in metropolitan areas where a smaller-sized CT with a higher population (on a fringe of a central city) neighbors a larger-sized CT with a smaller population (in a suburban area). That said, the latter one may have an abnormally high share of the studied group while being sparsely populated. The bivariate analysis allows for the non-accounting of such outlying cases when identifying areas of concentration.

The analysis of spatial correlation was conducted using GeoDa (version 1.22), the specialized software developed by the team led by Anselin and his colleagues [47]. This program also allows for the performing of an instant quantitative interpretation of the results obtained.

A shape file with spatial and statistical information on the census tract is imported into GeoDa, and then a matrix of spatial weights is made. In the current study, the geographical proximity between census tracts was determined based on the queen contiguity rule, which means that census tracts with a shared border or an angular point were considered neighboring CTs. The queen contiguity rule is well suited for a census tract grid with many right angles.

After the calculation of Local Moran’s I in this software, the LISA (Local Indicator Spatial Analysis) and the analysis of local indicators of spatial association were conducted. Every census tract that had at least one neighbor and whose Local Moran’s I value exceeded |0.05| was attributed to one of the following four clusters:

• HH (High–High) CT with a relatively high value of the independent variable neighboring a CT with a relatively high value of the dependent variable. Spatial autocorrelation is positive. These were designated as areas of concentration.
• HL (High–Low) CT with a relatively high value of the independent variable neighboring a CT with a relatively low value of the dependent variable. Spatial autocorrelation is negative.
• LL (Low–Low) CT with a relatively low value of the independent variable neighboring a CT with relatively low values of the dependent variable. Spatial autocorrelation is positive.
• LH (Low–High) CT with a relatively low value of the independent variable neighboring a CT with relatively high values of the dependent variable. Spatial autocorrelation is negative.

2.2. Assessing the Socio-Economic Status

To assess the socio-economic status of the concentration areas, we used the socio-economic status (SES) concept. This term has two basic definitions suggested by Villalba [48]:

• A construct reflecting the social and economic background of an individual or a group.
• A concept of a relative position in a certain social structure based on the assumption of a universally uneven status.

This study builds upon the second option. If we interpret it in terms of qualitative assessment, then this approach requires the comparison of community parameters with the mean values related to the overall structure. As of this study, the structure is the urban agglomeration, or to be more precise, its substitute—a metropolitan area, which is a totality of counties interconnected through economic and migration processes.

It is important to note here that it is not possible to single out “cognizant” communities, that is, social groups that are perceived (including spatially) by its representatives. In other words, we cannot identify vernacular districts and matching communities, while it is these categories (in a more mundane sense) that city dwellers operate in their daily activities. Therefore, socio-economic status will be assessed for people who “randomly” found themselves to be on the premises of one territory—a census tract.

The socio-economic status may be assessed using different sets of indicators. Those include socialization parameters, the quality of housing and environment (including access to services), accumulation of capital (poverty) and income level, the level of education, employment, and job qualification. If we narrow this diversity down to specific categories, three of them are being assessed most frequently: income, education, and employment [49,50,51]. The most inclusive set of categories is the one that was used by Darden and his co-authors to create the Darden—Kamel Composite Socio-Economic Index [52,53,54]:

• Median household income, US dollars.
• Unemployment rate among the working population, %.
• Share of the population with incomes below the poverty level, %.
• The share of people who completed higher education (with a bachelor’s degree) among citizens over 25 years old, %.
• The share of those occupying managerial positions or employed in areas of intellectual work among the employed, %.
• Median house price, US dollars.
• Median cost of renting a house (including operating costs and utility bills), US dollars.
• Share of occupied (not abandoned) houses, %.
• Share of households with a car, %.

For each territorial entity, the index is calculated as a sum of standardized values (z-score) of the above-mentioned indicators. Its explicit formula is missing in Darden and his co-authors’ publications, but it may easily be reconstructed as follows:

$${CSEI}_i=\sum _{i=0}^n{\displaystyle \frac{x_i-X}{S_{x_i}}},$$

where x_i is the x value in census tract i, X is the mean x value, and S is the standard mean deviation. The resulting values may be both positive and negative, which reflect, respectively, the higher and the lower SES.

Most of the studies devoted to the analysis of the SES of U.S. city districts use either one indicator (in most cases—the income level) or two to three indicators (poverty level and real estate value) [55,56]. The Darden–Kamel Composite Socio-Economic Index allows for the assessing of both the socio-economic characteristics (components 1, 2, 3, 6, 7, 8, 9) and institutional parameters (4, 5). Apart from that, some indicators pertain to society (1, 2, 3, 4, 5, 9), while others refer to the characteristics of the district itself (6, 7, 8). The reasoning of such diverse indicators was not found in other relevant studies. Finally, unlike classical indices, this assumes any values, both positive and negative, rather than the values 0–1. This is considered optimal when calculating deviations from the norm, as the index of 0–1 value may even out vast disproportions that exist among the districts of American cities.

3. Results

3.1. Concentration on State and County Levels

In terms of the aggregate population size, the two groups under study are very similar. As of 2018, the U.S. had 1.02 million Haitians and 1.12 million Jamaicans. They live across all the country’s states, but in some of them, their number does not exceed one thousand people, while other states include considerable concentrated shares of these groups (see Figure 1 and Figure 2).

Almost half of all Haitians live in Florida, which is unsurprising. This state’s coast is located closer to Haiti than other U.S. territories, and political instability in the country has repeatedly led to large-scale waves of Haitian refugee migration [14]. Later, the diaspora grew in the logic of chain migration. It should be noted that other states with the highest numbers of Haitian populations (New York, Massachusetts, New Jersey, Georgia, Pennsylvania, and Connecticut) also have access to the Atlantic Ocean.

The immigration of Jamaicans went more smoothly than that of Haitians, without abrupt surges similar to those in the 1970s and 1980s. The sharp influx of Haitians during this period was due to tens of thousands of refugees from the brutal political repression of the Duvalier family regime and later from the outbreak of civil war in Haiti. In this case, New York and Florida are the leaders, but the Jamaican population is distributed among them almost evenly. The aggregate number of Jamaicans in these two states also exceeds half of the population of the entire group in the U.S.A. The differences in the composition of other leading states in the Haitian case are minimal; on the list, we see Maryland instead of Massachusetts. Once again, all the states have access to the Atlantic. The top 10 counties by Haitian and Jamaican populations are shown in

Table 1. Counties with the largest Haitian and Jamaican population. Compiled by the authors based on [43].

County	Haitian Pop., Thousand	County	Jamaican Pop., Thousand
Broward, FL	126.8	Broward, FL	114.4
Miami-Dade, FL	126.3	Kings, NY	84.1
Palm Beach, FL	85.2	Queens, NY	65.8
Kings, NY	75.6	Bronx, NY	65.2
Orange, FL	49.4	Miami-Dade, FL	37.0
Queens, NY	33.1	Palm Beach, FL	35.7
Essex, NJ	24.2	Hartford, CT	27.0
Suffolk, MA	24.1	Westchester, NY	26.9
Nassau, NY	23.3	Nassau, NY	22.2
Middlesex, MA	21.3	Orange, FL	21.3

.

3.2. Concentration on a Metropolitan Area Level

All counties from the two lists are united by a high degree of urbanization as they are located near boundaries of large metropolitan areas. In both cases, the leader is Broward County in South Florida, which is explained by the geographical proximity to the Caribbean Sea and an abundance of large cities. Almost on a par with it is Miami-Dade County, where the principal international immigration hub of Florida and almost the entire American South—the city of Miami—is located. This county is Broward’s south neighbor. Overall, as was the case with the states, the groups of leading counties are hardly differentiated—the difference is only in the ranking position. Both lists include Florida’s county of Palm Beach (together with Broward and Miami-Dade, this makes up the Miami metro area) and Orange Country, the New York boroughs of Kings (Brooklyn) and Queens (both part of New York city), and Nassau County (part of the New York metro area). Haitians are widely represented in Essex County (the state of New Jersey, in the New York metro area), while Jamaicans are in Westchester County (which is also in the New York metro area).

This means that a considerable share of Haitians and Jamaicans is concentrated in two metro areas—New York and Miami. Indeed, the New York metro area has 235.7 k Haitians (23% of all U.S. Haitians) and 348.2 k Jamaicans (31.2% of all U.S. Jamaicans), while the Miami agglomeration has 338.2 k (33%) and 187.1 k (16.7%), respectively. In connection to such a substantial concentration of Haitians and Jamaicans in two metro areas, as well as in view of the large number of African Americans in them, the focus on the further analysis of the SESs of these two groups will be within the boundaries of the New York and Miami metro areas.

Three types of concentration can be singled out: only Haitians, only Jamaicans, and the two groups simultaneously. Apart from that, to draw a comparison with the African American districts, we identified the areas of concentration of African Americans that do not overlap with the concentrations of Haitians and Jamaicans. As the analysis of the spatial concentration implies, when identifying local extremums not based on the threshold values of population shares, the shares of titular groups in all four areas of concentration should be considered (see

Table 2. Shares of titular ethnic groups in the identified areas of concentration, %. Compiled by the authors based on [43].

Metropolitan Area	Haitians	Jamaicans	Haitian and Jamaicans	African Americans (Without Haitians and Jamaicans)
New York (mean) Miami (mean)	11.6	16.3	27.2	62.6
	18.5	6.6	27.4	63.5
New York (max) Miami (max)	29.4	45.8	55.2	100
	62.1	16.2	63.2	96.4
New York (min) Miami (min)	1	1.9	5.9	21.2
	1.4	1.8	7.1	19.8

).

As the analysis was performed using two variables, the areas of concentration included large CTs where the number of representatives of the groups under study is high, while the share is very low. In New York, the population of the CTs with 1% Haitians reaches 100 people, while in Miami, the population of the CTs with 1.4% Haitians is 102 people. That is, the aggregate population in such census tracts is almost 10k people (with the mean population density of CTs across the U.S. being 4.4 k people). It is noteworthy though that such cases are few (around 1% of the entire selection of census tracts that were among the areas of concentration). Also substantial is that both metropolitan areas had census tracts that were attributed to the areas of concentration of Haitians and Jamaicans, but were not attributed to African Americans. Figure 3 shows the areas of concentration of Haitians and Jamaicans in the New York metro area that have been identified through the analysis of spatial correlation.

Two large mixed Haitian/Jamaican areas are located within the city boundaries of New York—in the Kings (widely known as Brooklyn) and Queens boroughs. In the latter case, the mixed part of the area “gravitates” toward the hamlet of Elmont. Both areas are predominantly Jamaican. The Nassau Country also has such a Jamaican-dominated area. The single group-dominated areas, in turn, are located in other parts of the metro area. Predominantly Jamaican areas are seen on the border of The Bronx and the city of Mount Vernon, and predominantly Haitian areas are seen in New Hempstead, NY, Hillcrest, NY, Spring Valley, NY, East Orange, NJ, and Roselle, NJ. In all, we can single out seven areas of concentration of the two groups. Figure 4 demonstrates the concentration of Haitians and Jamaicans in the Miami metro area.

As is the case with New York, the areas have distinct borders and are distanced from one another. There are four of them: one is predominantly Haitian (in the cities of Delray Beach and Ocean Beach) and three are mixed ones. While the mixed areas in New York are mostly Jamaican, in Miami, they are mostly Haitian. This is caused by the sequence of migration waves: the Jamaican community of New York is “older” than the Haitian one, while the situation in Miami is just the opposite. When writing “in Miami”, we need to make a caveat that we specifically mean the metro area, as the city of Miami is, geographically speaking, a periphery of its own metro area, its southern part. The city is not even dominant in terms of population: of the 6 million people in the metro area, only 451k reside in the city of Miami.

The detached nature of the established areas in Miami is an interesting phenomenon. While New York has distinct suburbia that start on the fringe of the central city, Florida coast is a continued strip of highly urbanized territories with no distinct barriers between one another [57,58,59]. Despite this, as well as the relative nature of city boundaries, every concentration area is located at a considerable distance from its neighbors. Only two northern areas will possibly merge in the coming decades.

In both cases, we see an active process of Haitian and Jamaican areas overlapping. This means that the Jamaican and Haitian communities do not always occupy isolated immigrant quarters, yet this option is present too, more often in the New York metro area [60]. Overall, the bulk of both the Haitian and Jamaican areas as well as the mixed areas overlap with the territories of low racial diversity, populated predominantly by African Americans.

3.3. Socio-Economic Status of the Areas of Concentration

To identify how different the areas of concentration of Haitians and Jamaicans are from those of African Americans in the socio-economic context, the results of the Composite Socio-Economic Index for four types of areas should be considered:

• Areas of concentration of Haitians.
• Areas of concentration of Jamaicans.
• Areas of concentration of Haitians and Jamaicans.
• Areas of concentration of African Americans alone.

The results of the calculation can be found in

Table 3. Darden–Kamel Composite Socio-Economic Index values in the areas of concentration of Haitians, Jamaicans, and African Americans. Compiled by the authors based on [43].

Metropolitan Area	Haitians	Jamaicans	Haitians and Jamaicans	African Americans Alone
New York (mean) Miami (mean)	−3.72	−4.72	−1.55	−4.7
	−4.29	−1.27	−3.47	−5.35
New York (max) Miami (max)	4.72	7.46	7.07	7.3
	11.06	9.01	8.0	0.53
New York (min) Miami (min)	−12.76	−14.56	−14.24	−21.7
	−23.51	−14.12	−15.98	−16.08

.

In all area types in both metro areas, the Composite Socio-Economic Index assumes negative values, which means that, in general, Haitians, Jamaicans, and African Americans populate underprivileged districts. However, the indicators are substantially differentiated. In both metro areas, “newer” migrant communities—the Haitians of New York and the Jamaicans of Miami—live in more prosperous areas than “older” communities and African Americans.

Mixed areas are of special interest. While Miami’s SES values for mixed areas are somewhat logical (as a mix of features of Haitian and Jamaican areas), New York has an unexpectedly high SES value for mixed areas. The share of African Americans is incredibly high—75.7% (which is 8–39 percentage points higher than in other identified types of areas in the two metro areas). This means that in this case, a high share of African Americans is not connected with socio-economic ill-being. Moreover, these areas are more prosperous than suburban ones, which is untypical in the U.S. This may testify to the fact that the suburbanization of Haitians and Jamaicans follows a scenario that is different from the suburbanization of the white population as the SES of these two groups does not improve when they move to the suburbs. In metro Miami, the situation is reversed once again. The most prosperous areas are Jamaican-concentrated areas adjacent to mixed areas outside the central city of the metro area.

4. Discussion

Settlement patterns of Haitians and Jamaicans simultaneously have notable similarities and differences from those of African Americans alone. In a sense, the settlement of Haitians and Jamaicans is a model of how African Americans could have settled if the U.S. history had not had a long period of slavery and, consequentially, extreme spatial concentration [61]. Having passed by that stage of involuntary settlement, Haitians and Jamaicans, just like African Americans, have been concentrated heavily in central cities of major metropolitan areas (that said, they are virtually not present in the Black Belt). The New York and Miami metro areas are by far the most populated ones in terms of the size of these two groups [62].

Akin to African Americans, Haitians and Jamaicans in the New York metro area are actively settling in suburbia [63]. That said, this type of suburbanization may be considered “not exactly real” or even “fake”, as moving to the suburbs does not lead to a drastic change in status level, which is the case under the “classical” scenario of white population suburbanization. In the Miami metro area, the situation is compounded by the fact that there is no suburbia there in the traditional American sense [64]—the local settlement structure is formed by a narrow strip of cities and towns with no dominant ones (paradoxically, Miami is a periphery of its metro area), so instead of the suburbanization of Haitians and Jamaicans, we can rather see the formation of several detached areas.

There is a critical difference between Haitians, Jamaicans, and African Americans in factors forming SES levels. When African Caribbeans move to the U.S., they find themselves in an international setting of interracial relations instead of carrying the weight of discrimination that existed (and still exists in a structural way) throughout the history of the USA. “Newer” groups—the Haitians of New York and the Jamaicans of Miami—reside in much more prosperous neighborhoods as compared to the Jamaicans of New York and the Haitians of Miami who settled en masse in the second half of the 20th century when race-biased real estate market discrimination was a much more acute problem than it is today [65]. Similarly, less privileged are African American neighborhoods, most of which emerged on the sites of the old ghettos of the segregation age.

5. Conclusions

The study results in a conclusion that Haitians and Jamaicans are the “new type” of non-white ethnoracial groups. Unburdened by settlement structure flaws and a centuries-long legacy of socio-economic problems, they find themselves nonetheless among the underprivileged social strata, which sets them apart from many other migrant groups. However, as compared to African Americans living in the U.S. for several generations, they have fewer problems in terms of integration into labor market and the improvement of SES, as they are not loaded with the negative social capital accumulated by African Americans in the age of slavery and segregation.

The results of our analysis contributed to a rich body of work on the settlement patterns of ethnic and racial minorities in the U.S. and on the inequality between non-white Americans and the average level of socio-economic status. The novelty of our approach is to consider two large African Caribbean groups not as a part of African American community per sé, but as reference groups that are being analyzed in comparison with African Americans and the average level of socio-economic status in the selected cities.

However, our methodology has some evident limitations. First, we assume that African Americans are a somewhat uniform group in terms of their experience with inequality and systematic racism in different contexts. This is not exactly true because of the internal differentiation of African Americans as well as the variety of their socio-economic and geographic history. For example, some African Americans were victims of discriminative practices in the U.S. for several generations, while others could have lived in Africa or Antilles before reaching the U.S. and may have had to deal with other discriminative and violent policies of the colonial rule.

Second, we did not have chance to make a deeper analysis of the discriminatory policies that shaped the settlement patterns of Haitians and Jamaicans. This is very important for possible future research because Caribbeans have experienced overlapping patterns of discrimination, being both a non-white minority and migrants, but, in most cases, without becoming formally segregated, because there were relatively small African Caribbean communities in the Jim Crow South.

Another hypothetical change related to the statistical base could strengthen the predictive power of studies like ours. In addition to information on the origins of U.S. residents, one could examine the approximate time of arrival of their ancestors in the country. This would allow us to move away from analyzing static patterns to mobile ones and to testing in more detail the hypotheses that later migrations into the country may reduce inequality due to the absence of accumulated negative institutional capital.