In recent decades, immigration to New York City from Latin America has grown significantly both numerically and in diversity of source countries. 'Hispanic' neighborhoods have become nation-specific or region-specific enclaves which, in many ways, replicate cultural, social, economic, and racial stratifications found in Latin America. This paper examines the varying degrees of integration and segregation by national origin among Latin Americans in New York City as captured by the 1990 census. Through interviews with 87 Latin American immigrants and 32 real estate agents, we offer explanations for these patterns including preference for self-segregation, housing discrimination between Latin American communities, and the role of real-estate agents as possible 'gatekeepers' for particular neighborhoods based on national origin.

INTRODUCTION
Social scientists in the United States commonly refer to all immigrants from Latin America as 'Hispanics,' assuming that each of the source countries are sufficiently similar to support this broad category. Underlying this assumption is an unspoken expectation that historical conflicts, differences, and prejudices between source countries somehow disappear upon immigration. New York City, a primary destination for Latin American immigrants, serves as an excellent laboratory for examining the degree to which national loyalties are transferred to the United States and are evidenced in their settlement geography.

Hispanics from the Caribbean, Central and South America play a major role in recent immigration to New York City. Nearly one-quarter (23.3%) of the 1990-enumerated population of New York City self-identified as being of Hispanic origin (Table 1). The Puerto Rican and Dominican populations have historically dominated, with a combined population of 1.2 million. Central and South Ameriican populations, though smaller, are growing rapidly through family reunification (Table 2).

The purpose of this study is to examine the settlement geography of New York City's Hispanic population in order to identify the factors involved in the development of nationally based neighborhoods versus 'integrated' Hispanic neighborhoods. In the context of this study, integration and segregation refer to co-residence with, or separation from, Hispanics from other source areas. We explore national and linguistic preference and identity, race and class, urban structure (affordable housing, public transportation, ethnic economies), and the role of real estate agents as 'gatekeepers' in the settlement process. [end p. 31]

IMMIGRANT SURVIVAL STRATEGIES
The need to adjust to the demands of a new country leads many immigrants to co-reside in enclaves. The enclave offers a similar culture, language and very often employment opportunities to the new arrival. Smith (1995), Walker and Hannan (1989), and Loo and Mar (1982) argue that enclaves provide areas where immigrant language is spoken, customs and traditions practiced, and employment opportunities are available. They assert that enclaves are necessary for the survival of recent immigrants.

This is true of both immigrants and transnational migrants planning to return to their native country after what they assume will be a brief sojourn. In many cases the brief sojourn becomes a somewhat lengthy stay in the United States (Margolis 1994; Grasmuck and Pessar 1991). Settling in the immiigrant enclave becomes more than just a viable alternative to investigating the local real estate and job markets; it is their best means of survival in the United States. Settling in the enclave also enables transnational migrants to accomplish their goals more easily, which for many is to save enough money to establish themselves in their native country.

Many of these 'sojourners' may be in the United States legally upon arrival, but many overstay their visas. In addition to sojourners or transnationals, undocumented aliens also constitute a portion of those inhabiting and working in the ethnic/immigrant enclave. Undocumented aliens are prevalent among Hispanic immigrants in the United States. With business owners from their native country operating in the enclave, undocumented immigrants and visa overstayers can secure employment without legal working papers. They may even be able to secure a 'green card' from their bosses, something that would most likely not occur outside the enclave, especially with a limited knowledge of English (Margolis 1994).

Steering
Steering, the process whereby real estate agents direct clients toward certain areas due to the clients' racial or ethnic background, is illegal and kept in check through government monitoring. Even though steering is illegal, real estate agents have commonly practiced the process and have openly steered [end p. 32] prospective home buyers to areas where the racial background was similar to that of the client. The practice remains important today, however covertly undertaken. Palm (1986) suggests that race does affect real estate sales as evidenced by the actions of home sellers, home buyers and agents, all of whom tend to deal with their own races.

One method used to gather evidence of steering is to have auditors of different races play the role of prospective home buyers in order to analyze racial bias in locations shown by real estate agents. Galster (1990) found that race plays a monumental role in the real estate market. White teams were shown more houses and locations than black teams and were steered to locations that were white or which had a predominantly white school district. Blacks, on the other hand, were shown areas where people of African descent comprised a substantial portion of the population. The main difference between the two groups was the propensity for blacks to be shown areas that were adjacent to predominantly black neighborhoods. This kind of behavior clearly illustrates how real estate agents can be a driving force in the development of racial or ethnic enclaves. The influence of the real estate agent is also evident in Teixeira's (1995) study of Portuguese communities in Toronto. Portuguese agents were more likely than non-Portuguese to steer the immigrants to neighborhoods that were predominantly populated by those of Portuguese descent. Here, too, the agent acted as a catalyst for the continued growth of an enclave.

Integration in the housing market is different for Hispanic immigrants as a group than other immigrant groups. Hispanic immigrants are from many different countries with many different cultures, but more immportantly they are also from different races. Racially, Hispanics can be White, Black, Indian, Asian or any combination thereof. Racism still pervades American society and the real estate market in particular. This accounts, in part, for the differences in settlement patterns among Hispanic groups. Denton and Massey (1988) detailed the segregation between Anglos, Hispanics, foreign-born Hispanics, Blacks and Asians. They found more segregation between Anglos and Hispanics than between the two Hispanic groups and still further segregation between Anglos and foreign-born Hispanics. Denton and Massey (1989) also showed how racism affects the Hispanic immigrant community internally. Immigrants who self-identify as White tend to segregate themselves from those who self-identify as Hispanic.

The process of steering clients to certain locaations by real estate agents is a product of racism and ethnocentrism. Steering helps to create and develop neighborhoods based upon race or ethnicity and sometimes both. As a neighborhood changes from one ethnic group to another, nearby areas are often seen by real estate agents as transitional. When an area is seen as transitional, real estate agents steer clientele of the same ethnic background as those of the adjacent enclave to the area, thus expanding the enclave.

RESEARCH METHODOLOGY
A series of methodologies were used to examine the geography of Hispanics in New York City. First, 1990 Census data were mapped by census tract to show Hispanic settlement patterns by national origin in New York City. These maps reflect the varying patterns of clustering or dispersion of Hispanics by national origin. Entropy indices of diversity were computed and mapped at the census tract level in order to identify neighborhoods that were integrated and those that were segregated.

Second, interviews with 87 Hispanic immigrants were conducted between January and April, 1996, to discern factors motivating their housing choices. The respondents came from various ethnic and racial backgrounds, as well as from all five boroughs. 'Snowball' sampling (Margolis 1994) was used to select respondents. This technique relies on the respondents' friends and acquaintances as new sources for respondents, and it was quite effective in obtaining interviews. People were more apt to respond when they had been introduced through a mutual friend. Respondents came from every Spanish-speaking country in the Americas except Chile, El Salvador, and Paraguay. Although not a representative sample (65% of the respondents were from South America, 26% from Meso-America, and 9% from Cuba or the Dominican Republic), the interviews yielded very interesting insights into the role of national cultures in housing selection. The respondents were asked questions concerning where they lived and why they chose that location. The informal, open-ended questions usually led to a discussion in which the respondents related more details of their lives, including those which involved housing and ethnicity.

A common response to questions concerning neighborhood preference was access to affordable housing and public transportation. To test the role of urban infrastructure on settlement patterns, housing size, median rents, and public transportation access were entered into regression equations with selected [end p. 33] populations as dependent variables. Field observations confirmed interpretations of the regression analysis.¹

The third methodology involved surveying 32 real estate agents in the borough of Queens. Queens was chosen because the map of the entropy index showed that borough to have the greatest degree of diversity in both total number of Hispanics and in their settlement patterns. The real estate survey was carried out over a three-week period (April IS-May 3, 1996). Real estate agents were selected based on three criteria: (1) they served Queens neighborhoods with substantial Hispanic populations as of the 1990 census; (2) they advertised in the Hispanic Yellow Pages or Spanish-language newspapers; and (3) they specified offering services in Spanish.

The survey consisted of 18 questions focusing on factors influencing housing choice for Hispanic immigrant clientele. They were conducted in either Spanish or English based on respondent preference. The response rate varied depending upon the section of Queens, on the ethnic background of the respondent, and in some cases on how "corporate" the real estate agency was. Native-born agents were more willing to participate in the survey, particularly in Astoria and Long Island City. Until recently, these two neighborhoods have been predominantly Greek and Italian, and have experienced a large influx of Latin American immigrants from Mexico and Central and South America during the past decade. Hispanic realtors were relatively amicable, but some feared immigration and/or city officials checking to see if the agency was practicing any illegal steering.

In most American cities, local newspapers have many listings for no-fee apartments from real estate brokers or from those who own the dwellings. This is rarely the case in New York City. Unlike most other American cities, 86.9 percent of New York's 2.3+ million housing units are renter-occupied. Real estate agents control the information concerning rentals as well as home sales, and a fee for their serrvices is required. The fee is usually quite expensive in Manhattan, it is typically 15 percent of the yearly rent of the dwelling, and in the other boroughs it commonly equals a month's rent. With such a large sum of money required initially and credit background checks performed on prospective tenants, many immigrants become channeled to more inexpensive neighborhoods where fees are cheaper and credit background checks are relaxed. Thus, real estate agents become 'gatekeepers' controlling the access into most neighborhoods. They are also aware of owner preferences, and while they may not practice overt steering, they may inadvertently practice it.

HISPANIC SETTLEMENT PATTERNS IN NEW YORK CITY
Mapping Hispanic settlement patterns by national origin illuminates the clustering of many groups. Some tracts or neighborhoods have concentrations of multiple Hispanic ethnic groups; however, many more host only one or two. What is evident from the maps is the proclivity of certain Hispanic ethnic groups to cluster in the same areas or census tracts. In Queens, there are sections with high perrcentages of Colombians, Ecuadorians and Peruvians in the same locations. In other locations, there are clusters of Guatemalans, Hondurans and Salvadorans. In Manhattan, Dominicans and Cubans are often found in the same neighborhoods.

Hispanic immigrants may segregate themselves both by nation of origin and race. Often there is stratification by race in the country of origin. This is compounded in the United States where racism has prevailed for centuries and has helped to perpetuate segregated neighborhoods. Race can often play a role in the locational development of an ethnic immigrant enclave with studies showing a great deal of separaation between White Hispanics and Black Hispanics in housing (Denton and Massey 1989).

Census data suggest that this has happened to several groups in New York City. In the Dominican Republic, the predominant racial background is that of African descent or mulatto. According to the 1990 Census, Dominican immigrants in New York City have settled in many areas that have few other Hispanic immigrant groups and where the U.S.-born population is Black. Because Dominicans are so numerous in New York City, they can be found throughout the five boroughs (Figure 1). Where they have clustered in Washington Heights, there is also a significant Cuban population (Figure 2). Recent Cuban immigration has tended to replicate the population of the island with 30 percent of the immigrants having a Black racial background. With high levels of Black Hispanic heads of households, as well as a Black population in the census tracts, Cubans may be segregating themselves ethnically and racially. Alba and Logan (1993) have concluded that Cubans are more likely to live in areas with a high Black population, suggesting that the newer waves of Cuban immigrants are more likely to be of African descent and thus are more likely to live in neighborrhoods with a larger Black population. Skop (1997, [end p. 34]

[end p. 35] 56) found that of the 5,056 Marielitos in New York City in 1990, 55 percent were non-white. However, despite shared regions of origin, historic source country conflicts are manifested in differences beetween settlement patterns of the aforementioned groups and Puerto Ricans who are concentrated in East or "Spanish" Harlem, the Bronx, the Lower East Side, and the Greenpoint/Williamsburg area of Brooklyn (Figure 3).

Surprisingly, Ecuadorians and Peruvians tend to settle in the same neighborhoods, despite their history of border conflicts. They are concentrated in the Jackson Heights/Corona/Elmhurst area of Queens, an area where the early Hispanic settlers were Cuban refugees followed by a large influx of Colombians (Figure 4). In addition to co-residence, Colombians, Ecuadorians, and Peruvians have developed a powerful 'pan-Andean' enclave economy (Figure 5). Central Americans and Mexicans are relative newcomers to New York City; their communities have grown since the 1990 census, yet are still somewhat diffuse in their settlement patterns (Figure 6).

In order to identify diversified Hispanic neighborhoods, a standardized entropy index, a measure of the diversity of a population, was computed for each census tract and mapped (Figure 7). The entropy index of diversity applies an ecological model to population settlement patterns, measuring the degree of concentration or diversification of component groups under study. The index does not reflect the size of the total population under study but rather the relative representation of each component in that total population (Patil and Taillie 1982). The formula for the entropy index is as follows (Plane and Rogerson 1994: 302):

where 'n' is the number of subgroups in the population (14 in this study); 'Pk' is the population of the subgroup (for example, Dominicans); and 'P' is the total population (in this case of Hispanics), in each geographic area (census tract) with a maximum value of In n. The standardized index, computed as H* = H/in n, results in a score between 0 and 1.0. Zero is interpreted as the concentration of only one commponent group, and 1 infers equal representation of all groups.

Areas of highest diversity appear to follow pattterns of subway and bus lines, particularly in Queens (Figure 8), where many neighborhoods offer relatively affordable rents. This suggests a strong relationship between urban infrastructure, in this case access to affordable housing near public transportaation, and diversity in settlement patterns.

Informal Discussions with Hispanic Immigrants
Informal discussions with eighty-seven Hispanic immigrants were conducted to elicit reasons for residential choice. Table 3 lists the countries of origin of the respondents. As discussed earlier, respondents were identified via 'snowball' sampling. Although not a representative sample, these responses offer important insights into residential choices by national origin. Respondents answered differently based on national origin and seemed to reflect variations in socioeconomic and cultural elitism. Argentines resisted being called 'Hispanic' and seemed to be offended that they are enumerated as such. They considered themselves 'White' and would recount their European ethnic backgrounds. One Argentine interviewee called himself Italian, even though he spoke with a Spanish accent. Argentines repeatedly argued that their country was similar to the United States in that it was a country populated by European immigrants.

Responses by Argentines suggested a certain degree of prejudice and elitism. Many lived in neighhborhoods where the majority of the population was non-Hispanic White. Most agreed that there were certain ethnic groups, such as Dominicans and Puerto Ricans, with whom they would not live. Since approximately 85 percent of Argentina's population is white (Goodwin, 1998), it is not surprising that Argentine immigrants would prefer white neighbors. [end p. 36]

[end p. 37] Argentines were not alone in their attitudes. A Colombian woman stated that "Dominicans and Puerto Ricans are on public assistance," lowering the stature of the Hispanics in the United States. Another Colombian immigrant said he left his neighborhood because "there are too many Dominicans there now." Like Argentines, Venezuelans seemed to dislike being singled out as Hispanic. They perceived the term to represent someone from the Caribbean islands. Most Venezuelans interviewed lived in predominantly non-Hispanic white neighborhoods.

In speaking with Hispanic immigrants of various backgrounds, there seemed to be an expressed stratification by national or regional origin. South Americans were at the higher end, Mexican and Caribbean groups at the lower end, and Central Americans fell somewhere in between. Interviews conducted with other Hispanic groups showed less racism and elitism. Two Ecuadorian respondents disliked how Puerto Ricans spoke Spanish and said that they would not live in a neighborhood that was entirely Puerto Rican. The remaining three Ecuadorian respondents expressed little or no prejudice toward other ethnic or racial groups in terms of neighborhood selection. The motivation for choosing housing was principally economic--affordable housing in a safe neighborhood.

The Central Americans and Dominicans interviewed seem to care very little about race. They, especially Dominicans, considered themselves from a mixed racial and ethnic background and seemed less resistant to co-residing with other groups. Dominicans and Central Americans also stressed that housing affordability was the most important factor. They wanted a good neighborhood that was safe and not expensive. Living with people from their native country was important, but not as important as the price of an apartment or house.

Most of the respondents had been living in the United States for at least two years and had a good knowledge of the English language. Culture and language became less of a barrier in housing selection. For some, race and ethnicity was an issue, but the most important factors were affordability and safety. However, several respondents were recent arrivals who had settled in Hispanic immigrant enclaves. One such case was an Ecuadorian immigrant who chose to live in Corona, Queens. He chose Corona because he was able to secure a job in a restaurant and also because Ecuadorian friends helped him to find a roommate from Ecuador. Both these factors helped his adjustment to the United States.

Another case of a recent immigrant gravitating to an Hispanic immigrant enclave was a Peruvian woman who settled in the Bronx with her mother. She settled in a multi-ethnic Hispanic immigrant enclave in the Bronx with residents from Ecuador, Honduras, Mexico, Peru and Puerto Rico. Her choice in residential location was motivated by proximity to family and affordability. She said, "we can afford it [the rent]," unlike other places they had seen listed.

A recent immigrant from Argentina lived with relatives in Bellerose, an area of Queens that is not predominantly Hispanic, but traveled to work in the Hispanic immigrant enclave of Jackson Heights. Although he chose not to live there, the enclave provided employment and social and medical services not available in Bellerose.

THE ROLE OF URBAN STRUCTURE
Numerous respondents indicated that housing affordability played a major role in neighborhood selection. Housing size and costs vary drastically throughout New York City. Anecdotally, it was noted that many immigrants live in apartments with less than four rooms (studio or one-bedroom apartments). These smaller apartments account for 55 percent of the 2.3 million housing units in the city. Median rents vary more by neighborhood than by unit size. Studio apartments in Manhattan's Upper East Side can cost more than two-bedroom apartments in most Queens neighborhoods. Since only 30 percent of New York's population owns a car, access to public transportation also plays a central role in housing decisions, particularly for immigrants.

Regression analysis was conducted to assess the roles of housing size, rent, and access to public transportation in Hispanic residential patterns. Three independent variables were chosen: percent of housing units with fewer than four rooms per census tract, [end p. 38] median rent per census tract, and the absence or presence of a subway line in the tract (0/1 dichotomous nominal variable). The analysis was conducted for three populations per tract: the total number of Colombians, Peruvians, and Ecuadorians; the total number of Central Americans; and the normalized entropy index. The normalized entropy index represents multinational Hispanic neighborhoods. The analyses were conducted at two geographic scales--citywide (2,216 census tracts) and for Queens (672 census tracts).

Table 4 lists the standardized coefficients for the analyses. There is significant variation both spatially and by national group. Housing and public transportation explained very little of the variance in either South American or Central American settlement patterns citywide (adjusted r² = .058 and .059 respectively). In each case, the percent of smaller apartments in a tract was the most important of the three variables, with access to subways and median rent per tract playing much smaller roles. These variables explained more of the variance in Queens. Again, the percentage of smaller apartments was the most important variable and, for the South Americans, proximity to the subway played a significant secondary role. Low or negative values for rent reinforce statements made by respondents concerning housing affordability.

The normalized entropy index provides a surrogate measure of integrated Hispanic neighborhoods. When the housing and transportation variables were used to assess the variance in index patterns, the results were markedly different for both the entire city and for Queens. Availability of smaller apartments remained the most important variable, but median rent played a strong positive role. The results infer that higher diversity occurs in neighborhoods with higher rents, suggesting the importance of class factors such as higher incomes. Socioeconomic homogeneity may supercede national identity in residential preferences.

Proximity to the subway may not directly influence housing decisions. However, the pan-Andean enclave economy in Jackson Heights is centered on two 'ethnic main streets,' Roosevelt Avenue and Broadway, where the #7 and R trains intersect. Access to Spanish-speaking entrepreneurs often serves as a magnet for immigrants from all Latin American source countries.

Survey of Real Estate Agents
The real estate agent is often the gatekeeper of the community, especially in New York City. Housing sales and rentals often pass through the hands of the real estate agents working in the community. They make a decision of whether to show a [end p. 39] certain area to a client. Clients may come in and request housing in a particular neighborhood, trusting that they will be shown all possible listings. If nothing suitable is found, the real estate agent will often suggest an available listing. The suggestion may be linked to steering. Thirty-two real estate agents were surveyed in Queens in order to ascertain whether ethnic and/or racial steering occurred with Hispanic immigrant clients, as well as to determine if Hispanic immigrants were self-segregating based on national origin.

The realtors interviewed came from many different racial and ethnic backgrounds, but all offered services in Spanish and advertised in Spanish-language publications. Agents from neighborhood-based real estate offices were the most open and seemed to show a strong interest in the subject. Local offices of large corporate agencies tended to refuse to participate in the survey. Guyanese immigrants, who dominate the real estate market in southern Queens, were less apt to respond to the survey. They cited various reasons, but the most common was that they were too busy. Although most of the 32 agents interrviewed responded without hesitation, any question that seemed connected with steering was greeted with: "Steering is illegal. We don't steer our clients." Several questions that were more specific, especially dealing with the race of Hispanic immigrant clients, were more difficult and needed further explaining. The most thorough responses were obtained from Hispanic realtors, who seemed to understand the nature of the study more easily than non-Hispanic realtors (see Figure 9).

The real estate agent may not be the principal party in steering; often the process begins with the client. In many cases, those selling houses or renting apartments will only consider dealing with certain ethnic groups. This is not only the case with White Americans, but with many Hispanic ethnic groups as well. Several respondents in Woodside and Sunnyside admitted to clients with certain preferences. One informant said, "We rent a lot of apartments to Spanish people, but most of our clients don't want Mexicans. They feel they [the Mexicans] will try and share the apartments with too many other people." Another agent two blocks away said, "We sell a lot of houses. Our clients don't want to sell to Dominicans or Puerto Ricans." In the same neighborhood a different respondent stated that" Argentines only want to sell to Argentines." This reference to Argentines occurred repeatedly.

Most studies on steering help to illuminate the problem on a racial level (Mohl 1995, Wiese 1993, Galster 1990). Agencies and sometimes real estate developers steer clients to certain areas based on skin color. They discuss block-busting slumlords and those looking to change a community for personal gain. However, most of the steering committed by real estate agents found by this study was inadvertent and in many cases assumed to be beneficial. The agents simply respond to the desire of the building owners. The agents know the owner will only sell or rent to certain people. To avoid wasting the time of all those concerned, they withhold information concerning certain places for some clients. The most frequently cited group of Hispanic immigrant owners not wanting to sell or rent to others was that of the Argentines. Several agents said they had buildings owned by Argentines and these owners preferred Argentines as tenants. Accepting others was more difficult unless the client seemed to fit the stereotype of a tenant the Argentines wanted. Otherwise, showing the apartment was not worthwhile.

Housing discrimination is not confined to those on the selling or renting end, but also is practiced by those seeking housing. At times the clients will specifically ask for certain neighborhoods based on their own racial background. One respondent in southern Queens concluded, "If they [Hispanic immigrants] have light skin they usually prefer White neighborhoods. If it is darker they usually don't care." When the same respondent was asked as to whether this pertained more so to any particular Hispanic ethnic groups she said, "Yes, usually with the Colombians." Another informant stated that this was also the case at his agency. The day of the interview, a Caribbean Hispanic was shown a house in a Black neighborhood. The real estate agent suggested this area due to the client's physical features, but when the client saw the area he said, "This is a Black area. I don't want to live in a Black area. I'm not Black. I'm Puerto Rican."

Dominicans appear to bear the brunt of the discrimination between Hispanic ethnic groups. Sentiments often expressed about the Dominicans were that they were too loud and too friendly. Many also said that Dominicans were on public assistance more often than other groups. Whether it is due to culture, color, or economic background, Dominicans appear to be the Hispanic immigrant group that faces the most discrimination, despite their position as the second largest Hispanic group in New York City. Dominicans also appear to be discriminated against by other racial and ethnic groups. One informant, a [end p. 40] Dominican, stated that Koreans and Chinese were hostile and suspicious of Dominicans when doing business with them. This real estate agent said he had handled several housing sales for Asian clients who dismissed dark-skinned clients brought to them.

Although steering is illegal, many real estate agents practice it to varying degrees, as evidenced by responses to a question on whether the client usually chooses an area independent of the agent's suggestion. They claimed that everyone entering their agencies knew exactly where they wanted to live when they walked in the door. Many respondents stated that steering was illegal and they never steer anyone. However, as interviews progressed, many of the described practices, although not exactly steering, showed evidence of housing discrimination by landlords.

Results from the interviews suggest that there is a higher propensity for certain groups to use the same real estate agency. Central Americans, particularly from El Salvador, Guatemala and Honduras, have been clients at the same agencies. South Americans showed a similar propensity, with agencies having clients from Colombia, Ecuador and Peru. Although few in number, agencies that had a Bolivian clientele also catered to Chileans and Paraguayans.

South American immigrants seemed to resist living near Caribbean Hispanics, and Dominicans were not interested in living with Puerto Ricans. Guatemalans, Hondurans and Salvadorans had the least animosity towards other Hispanic ethnic groups. Real estate agents stated that these ethnic groups were interested in the price of an apartment or a house, not in who was living in the neighborhood.

A large percentage of Hispanic immigrant clients preferred neighborhoods with people who shared their race, national origin, or culture. Approximately 19 percent of the agencies said that their clients preferred neighborhoods where there was a substantial population of the same national background. Racial similarity was important to some clients as well. Six percent of agencies said that their clients preferred areas of the same racial makeup as themselves. Another 15.6 percent of agencies had clients who preferred a neighborhood with a population of both the same national origin and race. This suggests that Hispanic immigrants may be segregating themselves racially as well as nationally.

In many instances, word of mouth helps to drive the real estate business. Sixty-five percent of the respondents said they regularly received new clients because of referrals from a friend or relative. With such a high percentage of referrals, the ground is laid for the real estate agent to help establish a nation-specific enclave. Larger agencies have multiple listings and can service a very large area. In New York City, though, many more, including over 60 percent of those interviewed, are smaller and mainly service the immediate area. With territory limited to a small area, the likelihood of a referral living in the immediate vicinity of the person(s) who referred them increases.

Although findings from other studies (Palm 1986) suggest that minority clients tend to be drawn to real estate agents of their own background, this did not necessarily hold true in the data of this study. The findings of the real estate survey show that Hispanic immigrants patronize agents of different races and backgrounds. Perhaps this is because New York City, and especially Queens, is so multiethnic that dealing with the same ethnic background is not always the easiest or best approach in seeking out housing.

Many real estate agents advertise in the Spanish-language newspapers with a large immigrant readership. Several of these advertisements list people to contact who do not have Spanish names. If this results in a referral to someone who is a native Spanish speaker, the process of steering ethnically or linguistically has begun. Even if a non-Hispanic helps the prospective client, the Hispanic community is being courted and the agent will most likely have in mind what will fit the clients' needs. As clients are directed to agents acquainted with their ethnic background and language, the agency has started the process of steering (Teixeira 1995).

One respondent, a Korean immigrant, said that all groups prefer to live with people from their country. His own attitude would play a part when dealing with clients. He had Mexican clients and said that they prefer to live near one another. One cannot say for certain how often his Mexican immigrant clientele preferred to live in the same neighborhood but, with his attitude, one can infer that he would show them places in a 'Mexican' neighborhood.

CONCLUSIONS
Scholars and non-scholars alike have become so accustomed to using the term 'Hispanic' that one forgets that it is an umbrella term for a very diverse population whose principal common characteristic is a shared history of Iberian colonization. Politicians and other social leaders presume that, upon arrival in the United States, Latin Americans will develop a pan-Hispanic identity and polity (Miyares 1993). [end p. 41] However, each Latin American country has its own distinct identity and may have a long-standing history of animosity or prejudice toward those of another Latin American country. It is naive to assume that these prejudices will disappear upon immigration to the United States. It is also naive to think that all Latin Americans will assume an 'Hispanic' identity. The retention of national origin identity is clearly evident in the settlement geography of Latin Ameriicans in New York City. Whether by choice or direction, several Latin American borders have been recreated in New York City.

Urban structure influences where immigrant neighborhoods form. In New York City, areas where affordable housing and public transportation coincide tend to become magnets for immigrants from all nations. This can be explained by enclave economies that have formed along major transportation corridors (in this case, subway lines). Jackson Heights, Elmhurst, and Corona have become focal points of the Hispanic economy in Queens and are multinational Latin American neighborhoods. Here, economic cooperation has superceded source country conflicts and has attracted a very diverse Hispanic population.

In New York City, real estate agents play a powerful gatekeeper role in explaining who settles where. They are the ones who control information concerning available listings, regardless of tenure type. This study has shown that agents facilitate segregation by national origin. This occurs overtly by steering or directing clients to neighborhoods populated by those with a shared national origin, race, or culture or by preventing access to places when the agents expect that the owners will reject the prospective renter. This also occurs inadvertently when agents assume clients would prefer to live in national origin-specific neighborhoods and show available listings there, intending only to serve the best interrests of the client and not to facilitate any form of segregation.

Race and class are powerful forces in Latin America and in the United States and they play themselves out in New York City. The more 'European' the country of origin, the more likely the immigrant will be to settle in a non-Hispanic White neighborhood. Dominicans, who are overwhelmingly of African descent, experience the double impact of being both Black and Hispanic and suffer prejudice from predominantly European Latin American immigrants as well as from non-Hispanics. Segregation by skin color is not solely a factor of immigrant status or whether one considers oneself to be Hispanic, as is implied by Denton and Massey (1988, 1989).

In New York City, Latin American borders have been recreated. Despite statements by respondents concerning national preferences in residential selection, the 'borders' do not seem to be national ones. Instead, distinctions by race and class, common throughout Latin America as well as the United States, have resulted in segregated settlement, often reflective of source-country patterns.

REFERENCES
Alba, R. D., and 1. R. Logan. 1993. "Minority proximity to whites in suburbs: An individual-level analysis of segregation," American Journal of Sociology 98(6): 1389-1427.

Denton, N. A., and D. S. Massey. 1988. "Residential segregation of Blacks, Hispanics, and Asians by socioeconomic status and generation," Social Science Quarterly 69(2): 797-817.

___ . 1989. "Racial identity Among Caribbean Hispanics: The effect of double minority status on residential segregation," American Sociological Review 54: 790-808.

Department of City Planning. 1996. The NEWEST New Yorkers 1990-1994: An Analysis of immigration to NYC in the early 1990s. New York: New York City Department of City Planning, December.

Galster, G. 1990. "Racial steering by real estate agents: Mechanism and motives." Review of Black Political Economy 19: 39-63.

Goodwin, P. B. 1998. Global Studies: Latin America. 8th edition. Guildford: Dushkin/McGraw Hill.

Grasmuck, S., and P. Pessar. 1991. Between Two Islands, Dominican International Migration. Berkeley, CA: University of California Press.

Loo, c., and D. Mar. 1982. "Desired residential mobility in a low income ethnic community: A case study of Chinatown," Journal of Social Issues 38(3): 95-106.

Margolis, M L. 1994. Little Brazil, An Ethnography of Brazilian Immigrants in New York City. Princeton, NJ: Princeton University Press.

Miyares, J. M. 1993. "Finding common ground: Unity in diversity in the political mobilization of U. S. Hispanics," Erdkunde 47: 229-237. [end p. 42]

Mohl, R. 1995. "Making the second ghetto in metroopolitan Miami, 1940-1960," Journal of Urban History 21(3): 395-427.

Patil, G. P., and C. Talie. 1982. "Diversity as a concept and its measurement," Journal of the American Statistical Association 77: 548-563.

Palm, R. 1986. "Racial and ethnic influences on real estate practices," The Social Science Journal 23(1): 43-53.

Plane, D. A., and P. A. Rogerson. 1994. The Geographical Analysis of Population With Applications to Planning and Business. New York: John Wiley.

Skop, E. H. 1997. Segmented Paths: The Geographic and Social Mobility of Mariel Exiles. Unpublished M.A. Thesis, Department of Geography, Arizona State University, Tempe, Arizona.

Smith, C. J. 1995. "Asian New York: The geography and politics of diversity." International Migration Review 29(1): 59-84.

Teixeira C. 1995 "Ethnicity, housing search and the role of the real estate agent: A study of Portuguese and Non-Portuguese real estate agents in Toronto," Professional Geographer 47(2): 176-183.

U.S. Census. 1990. Census of Population and Housing, STF 3A. CD-ROM version. Washington, DC: U.S. Bureau of the Census.

Walker, R., and M. Hannan. 1989. "Dynamic settlement processes: The case of US immigration," The Professional Geographer 41 (2): 172-82.

Wiese, A. 1993. "Places of Our Own: Suburban Black Towns Before 1960," Journal of Urban History 19: 30-54.

RESUMEN
Durante los últimos treinta años, inmigración de la América Latina hacia la ciudad de Nueva York ha crecido significativamente por cantidad y por diversidad. Los barrios hispanos se han formado a través del país y de la región de origen de los residentes. Los barrios replican de varias formas las estratificaciones culturales, sociales, y económicas que se encuentran en la América Latina. Este articulo se trata de los grados de integración y segregación por origen nacional que se encuentran en Nueva York, informado por el censo de 1990. A través de entrevistas con 87 inmigrantes latinoamericanos y 32 agencias inmobiliarias, ofrecemos un análisis de los barrios hispanos, incluyendo la preferencia de segregarse, la discriminación residencial entre las comunidades latinoamericanos y el papel del agente inmobiliario como 'guardia de puerta' de los barrios particulares a través de origen nacional. [end p. 43]