Edward L. Jackiewicz and Yifei Sun

., 1005, 5 (1-2)? 114 /J 8 The Ties that Bind, or Not? The Assimilation of Brazilian Migrants in South Florida Edward L. Jackiewicz and Yifei Sun This paper examines the immigration and assimilation experiences of Brazilians in South Florida. Brazilians have been largely overlooked in immigration literature. However, they are an increasingly important part of the changzng ethnic mosaic in large U.S. cities, particularly in South Florida. This research builds on the notion of Brazilians as the "invisible minority," put forth in Margolis' study of Brazilians in New York City. To achieve our goal we conducted surveys with Brazilians in South Florida to examine their immigration experiences and feelings about the local Brazilian community. Ourfindings show that the invisibility of the Brazilian community is largely a function of the desires of the Brazilians themselves who wish to integrate quickly and, at least initially, earn money and return home. Howeuer, as the immigrants extend their stay (at least partially due to the ongoing economic hardships at home) they become more integrated and feel more strongly about strengthening the Brazilian community, which in turn is likely to heighten their visibility. Keywords: Brazilian immigrants, South Florida, ethnic geography. Introduction This paper examines the immigration and assimilation of Brazilians into South Florida. South Florida (defined here as Miami-Dade, Broward, and Palm Beach counties) is a well-known receiving area of migrants from throughout the Caribbean and Latin America. The experiences of Haitians and Cubans in this region have received the greatest attention to date (see Portes and Stepick 1993, Stepick and Portes 1986, Boswell 1993, Boswell and Curtis 1984) but immigrants from many other countries also contribute to this region's rich ethnic mosaic. The stream of Brazilians into this area is one such group that is much less documented even though they represent an important element of the changing demographics of this area. The immigration of South Americans, including Brazilians, into the United States has received little attention from researchers and policy makers alike, yet since the 1980s sigruficant numbers arrived from this region. One reason that Brazilian immigration into the United States has received little attention is perhaps the absence (or at least perception) of a visible community in the areas of settlement. This has occurred to the point that Margolis (1998) refers to them as the "Invisible Minority" in her book about Brazilian immigration to New York City. Beginning with the work of Park and Burgess in 1920s Chicago, scholars have been interested in the assimilation of immigrants into urban areas. The early consensus was that as immigrants and their descendants improved their relative economic status they would leave their racially and ethnically homogeneous inner city enclaves to relocate and integrate into the more diverse areas away from the urban centers (Massey 1985). While this held some accuracy in analyzing the experiences of European immigrants, it is less relevant to the Latino and Caribbean experience, especially in

multiethnic areas such as South Florida. This begs the question: to what are they assimilating? This paper is unique in that it empirically examines the causes of this invisibility through survey research and quantitative and qualitative analysis. Additionally, this research examines the Brazilian immigration and assimilation experience in South Florida and how it complies with existing theories of assimilation. Within this framework, the emphasis is on the role of agency among these immigrants and how they are shaping their experience through personal decisions and, in turn, defining Brazilian assimilation. Thus, this research contributes to the wider literature of the immigrant experience as well as the less developed literature on the Brazilian diaspora. Brazilians on the Move Since the economic crisis of the 1980s, Brazilians increasingly tested the waters of international migration. The economic crisis in Brazil during the 1980s and extending into the 1990s placed unprecedented hardships on the country's citizens and emigration was a rational response. The 1980s, referred to as the "lost decade" in Latin America was characterized by low wages, underemployment, economic insecurity, and hyperinflation. Between 1982 and 1994, Brazil's annual inflation rate was never less than 98 percent and exceeded 1000 percent in five different years reaching a peak of 2,937 percent in 1990 (Franko 1999). There were also three years during this period when the economy had a negative economic growth rate and there was a wage freeze in 1986, which prompted the introduction of a new currency, the cruzado, to provide temporary relief. Within a year and a half of implementation of the plan, however, inflation again exceeded 1000 percent (Franko 1999). After much government tinkering in price freezes, wage freezes, currency devaluations and various other plans, another new currency was introduced in 1994 and by 1996 inflation had been stabilized and there was modest economic growth. However, this did not reduce the wide socioeconomic disparities in Brazil nor did it overshadow structural weaknesses in the economy as was evidenced when the "Asian Flu" spread to Latin America resulting in large capital outflows. In short, economic stability has been elusive in Brazil and it has crippled the daily existence of many Brazilians and provided a decisive "push" factor creating a- large stream of economic refugees. The number of Brazilians that actually fled the country and landed in the United States is almost impossible to estimate, mainly because so many are undocumented. For example, the U.S. census for 1990 estimated that there were 94,023 Brazilians living in the United States. In contrast, the Brazilian Ministry of External Affairs puts that figure for 1996 at approximately 600,000. Why such a wide disparity? In addition to the fact that so many are undocumented, many arrive in the United States on tourist visas, stay beyond their time limit and become undocumented (Margolis 1998). The INS has no specific programs to locate and deport the "overstayers." This does

not mean that there is no element of fear of deportation among the illegal residents. Thus Brazilians, similar to any other immigrant group residing in the U.S. illegally, are virtually impossible to count, largely because many do not want to be identified. Margolis (1998) adds many Brazilians view their stay here as temporary and have little interest in participating in a census. She also notes Brazilians do not fit neatly into standard U.S. ethnic or racial categories. They are not "Hispanic" because they speak Portuguese and many do not consider themselves to be either black or white. These factors contribute both to the inability to accurately tabulate the number of Brazilians and their "invisibility" (Margolis 1998,B-9). The reasons for moving to the United States are also driven by economics. Many Brazilians are fully aware of the wage differentials between the two countries and view migration as a means to improve their economic well-being. Many of these people can be classified as "sojourners" rather than migrants because they intend on staying in the United States for only a short period of time and plan on returning once a certain economic goal is attained (e.g, enough savings for a down payment on a home). This economic migration can be traced back to World War I1 and the mining town of Governador Valadares in the state of Minas Gerais. Americans had multiple links with this mining industry and after the war many hired Valadarenses as household servants and brought them back to the United States. This helped to create a culture of out-migration in this area and now it is estimated that 45,000 Valadarenses live in the United States and send back an estimated $12 million each month in remittances (de Souza 1996). More recently however, it seems that the sojourner philosophy is being replaced with one of permanency. Sales (1998) research on Brazilian immigrants in Boston shows that many Brazilians now desire to remain in the United States for longer periods than researchers previously estimated. While the reasons for these extended stays may vary, the prolonged economic crisis in Brazil is certainly a determining factor. Whether or not this new attitude will result in changing feelings toward the host society and the creation of a more visible community has yet to be evaluated. Goza (1994) in his study of Brazilian immigration to North America found that it was quite common to migrate as family units, but the difficulty of entering the country legally led to the breakup or separation of many family units. Many enter the United States through Mexico with the use of a "coyote." Many travel alone, but utilize established social networks in destination areas. Making It Work-Theories of Assimilation The most recent literature on immigration seems to concur on at least one point: the immigration experience of contemporary migrants, particularly those from Asia and Latin America, differs greatly from that of the European experience at the turn of the twentieth century (Portes and Rumbaut 1996).

The North American Geoguapher This is partially attributable to the wide diversity of immigrants since the relaxing of immigration laws in 1965. The backgrounds of contemporary immigrants differ in many regards, including education, skill-levels, economic and social resources, English fluency, and place of origin (notably rural vs. urban) all of which influence their ability to assimilate (Alba and Nee 1997). Additionally, many contemporary immigrants have different motivations than earlier immigrants who sought a new home. Contemporary immigrants often seek to earn higher wages and to send as much money as possible back home and have every intention of returning when they attain their financial goal. Of course, these intentions often change after an extended stay in the host society and the prospects of returning home seem less desirable. Although the recent immigrant experience in the United States differs widely from when Europe was the dominant source area, Alba and Nee (1997) argue that assimilation theories need to be amended rather than abandoned to better address the contemporary experience and shed its historical association with ethnocentrism and notions of Anglo conformity. In other words, the assimilation of immigrant communities may not in fact require them to become more "American," but rather to retain important elements of their national identity and transform the socioeconomic landscape of the host society. It is here that the role of agency, that is the desires and intentions of the immigrant, becomes central to the experience. Therefore, as Rumbaut (1997) suggests, if we are to approach assimilation from diverse perspectives and move beyond dichotomies, then a more detailed analysis of immigrant experiences is necessary. The growing realization that Asians and Latinos would not assimilate as fully into the U.S. social and economic systems as did Eastern Europeans at the turn of the century was the basis for segmented assimilation theory, attributable mostly to Portes and Zhou (1993). Segmented assimilation theory posits that exit conditions, class, and entry context will all significantly influence the assimilation experience and those entering under precarious circumstances, with little or no English skills, job skills or financial resources, and/or little or no institutions or individuals to help them initially will have a much more difficult time integrating and succeeding in the host society. Distinguishing immigrants in these various ways is useful in determining the nature of adaptation and propensity for economic success. It is also important to point out that assimilation, at least in the way it is commonly understood (e.g. learning the language and customs of the host society) is no longer as essential as it was a century ago, particularly when referring to Latin Americans in South Florida. As Gans (1997) argues, the range of immigrant experiences is often reduced to a dichotomy (i.e. successes and failures) when in fact it is a reconciliation between retentionism and acculturationism. Gans also adds that it

is essential to look at the voluntary and involuntary nature of retentionism as some immigrants fight hard to retain their cultural identity while others seek to shed it as quickly-as possible.. The absence of visible Brazilian communities throughout the United States is perhaps an indication that many of these immigrants are not primarily concerned with retaining an ethnic identity. While segmented assimilation focuses on individual attributes, the queuing theory of assimilation prioritizes group characteristics. Queuing theory purports that group membership takes priority over skill level and that workers stand in an imaginary line waiting for the good jobs and the place in line is determined by some measure of desirability (i.e. race, ethnicity, skills) and as such it factors in employer discrimination (Model and Lapido 1996). The large, successful Latino population in South Florida suggests that Brazilians are not likely to be discriminated against in this area. Upon arrival, Brazilians, like nearly all immigrants, confront many challenges such as finding housing and employment, learning the language, and coping with culture shock. However, some are more prepared and handle these challenges better than others. Ideally, earlier migrants such as friends or relatives might have housing or employment options lined up, -which can greatly ease the assimilation process, however this is often not the case. To find legal employment, immigrants need to acquire a social security card. In South Florida, as in other major immigrant areas, it is relatively easy to buy a counterfeit one for about $150 (Margolis 1998). This is the passport into the working world and a necessary first step for the new immigrant. Since many Brazilians, at least initially, do not view their stay as permanent, they do not search for an ideal living arrangement and often people will share space to save money. Also, Brazilians tend to be less bothered by this type of accommodation than would many Americans. In fact, many often choose to live with several people so as not to be lonely (Margolis 1998). Additionally, they have, little interest in strengthening the local Brazilian community and in turn, increasing their "visibility." According to Sales (1998), Brazilians have been typecast as hardworking and compassionate, attributes that have been helpful in their securing jobs in certain sectors, and many find jobs caring for the elderly, cleaning homes, working construction, dishwashing and office cleaning. Goza (1994) found in his research that regardless of gender the most common job was a janitor. He also found many men working as landscapers, dishwashers, or car washers, while women frequently worked as waitresses, babysitters, cooks, and housekeepers. Also, because Brazilians are often associated with Hispanics who are perceived as being lazy, they try to differentiate themselves and demonstrate a strong work ethic as quickly as possible. Most important is that the newly-arrived population is eager to make money to pay off debts or begin saving. Notably absent from this list of occupations is entrepreneurial

activities or small business owners. These occupations are common among many other immigrant groups and are often central to the establishment of an immigrant community and the visibility of that group. Many new arrivals come from middle-class backgrounds in Brazil and often take menial jobs, which can be a difficult transition and tough on the psyche. Under these circumstances, the familial gender roles may flip-flop and the woman who may have been a housewife in Brazil is now the largest breadwinner of the family and provides these women with unprecedented financial autonomy. This situation can cause stress within a marriage and lead to divorce or separation. Margolis (1998) found that many Brazilians change after arrival from caring, generous individuals to selfinterested money hoarders resulting in a lack of unity within the community. Similarly, group solidarity is limited due to the absence of a functioning community organization that aids in the settlement and assimilation of Brazilians as with other ethnic groups. This lack of social community is reflected in the absence of a physical community and hence the "invisibility." There tend to be very few ~razilian enterprises that would function as an important welcoming center for new arrivals. This is not to dismiss the fact that upon arrival many Brazilians are aided by their friends and relatives. As the assimilation process occurs many decide that they would rather not return home and many become transnational migrants maintaining active ties with family and friends in Brazil and make frequent visits, if they are able to return. Many, through remittances, have a constant financial commitment to Brazil through this transfer of money and other investments. The decision to stay longer than intended will likely continue as long as the relative financial situation remains better. Further, many.of the Brazilians who migrate are highly qualified and educated, and in some ways alleviate the problem of underemployment back home, while simultaneously providing a cheap supply of skilled labor to industrialized countries and fill jobs. that might otherwise remain vacant. To summarize, we hypothesize that the "invisibility" of the Brazilian community is mostly attributable to the desires of the immigrants themselves, but it is lessening over time as the Brazilian community becomes larger and more established. Data and Methodology The data for this study were collected through a survey that includes 29 questions divided into four categories. Interviewees were not preselected, but rather were identified through various means (e.g. churches, Brazilian grocery stores, friends/acquaintances) and the "snowballing" technique was utilized to expand the survey base. The four categories included in the survey are: personal information, information about the immigrants' status and livelihood in Brazil, information about their experiences arriving and settling in the United States, and other data which include

The North Americ!an Geographer feelings about returning to Brazil. Space was also reserved for additional comments the interviewee thought were not addressed through the survey. In total, we collected 45 surveys, although a few did not answer every question so the number of valid observations varies from 42 to 44 depending on the nature of the inquiry (Figure 1). The theorization of the research and the resultant survey questions draw on the more general literature on migration and assimilation and the limited research conducted on Brazilians in the United States. As such, the objectives of this study are two-fold: first, we seek to understand the factors influencing Brazilians' assimilation to the United States including aspects of their invisibility; second, we aim to examine the consequences of assimilation on the Brazilian immigrants (i.e. how assimilation or integration into the United States has altered their attitudes about returning home, becoming a U.S. citizen, and building ties within the Brazilian community). Assimilation operates on various levels and this study looks into political, social, and psychological aspects of assimilation which are measured through three dependent variables respectively: whether or not the person is a U.S. citizen (US-CITIZEN); the percentage of one's time spent with others from Brazil (ACT-WBRA); and whether or not the person intends to return to Brazil (RETURN). It is assumed that if a person has received U.S. citizenship, spends little time with other Brazilians, and has no intention to return to Brazil that the person has a high degree of assimilation. It is important to note that we assess the amount of time spent with other Brazilians in two ways. Initially, we test this with a continuous variable ACT-WBRA utilizing the actual percentage of time each respondent claims to spend with other Brazilians. A second evaluation of this variable is because many immigrants report that they spend more than 50 percent of their time with other Brazilians, therefore the requirements of a normal distribution to use regression models are not satisfied, we converted it into a binary variable, ACT-BRA50. There are several independent variables that affect the assimilation process considered in this study, including: age (AGE), education (EDUCATION), English proficiency prior to arrival and current English proficiency (ENG-P and ENG-C), gender (MALES) and years in the United States (YRS-INUS). It is hypothesized that older people may confront greater challenges assimilating than a younger person, all other factors being equal. Meanwhile, it is hypothesized that education, English proficiency and length of time in the United States are all positively correlated with assimilation. Regarding the impact of gender on assimilation, it is generally hypothesized that males are more likely to be further integrated into mainstream society primarily because they are most likely to be the primary wage-earners and spend a greater amount of time interacting with members of the host society. However, recent studies have found that there is often a role-reversal among immigrant families and females

Binary Variables ACT-WBRA50 Total valid observations 44 Frequency Value = 1 22 Percent 50 Frequency Value = 0 22 Percent 50 - AGE ACT-WBRA (Oh) Education Income (US $) I Minimum 1 0 0 1 Maximum Average YRS-INUS 0 30 8.04 6.92 I I I I I I Figure 1. Summary statistics of variables. 5 100 3 5 2.67 57.02 1.67 2.41 Standard deviation 1.17 31.73 0.60 1.27 are as likely as males to be in the workforce and in many cases earn more than their partners. As previously mentioned, this can also be a major source of conflict within the household as many males become uncomfortable with lessened status within the family. In this analysis, we also address one facet of the impact that assimilation may have on the local community, namely sentiments about making the Brazilian community stronger. This model uses the feelings about strengthening the Brazilian community (FEEL-BRACOM) as the dependent variable and ENG-C, EDUCATION, YRS-INUS, MALES, as control variables in addition to US-CITIZEN, RETURN, and ACVBRA50 as the independent variables in Models 1, 2, and 3 respectively (Figure 4). It is assumed that if many individuals want to strengthen the Brazilian community, then this will in turn lead to heightened visibility of Brazilians in general because there will be greater congregations of Brazilians and greater usage of public space. It is reasonable to expect that the more the immigrants are assimilated, the less likely they are to want the local Brazilian community to be stronger. However, it is argued here that the opposite will occur. As many Brazilians attain their personal goals and assimilate to a certain level, many will seek out other Brazilians for various reasons and effectively strengthen the Brazilian community.

I Variables AGE ACT-WBRA ACT-WRASO DISCRIMINATION EDUCATION - ENG-P ENG-C The North American Geographer Explanation and coding 1: <25; 2: 25-34; 3: 35-44; 4: 45-54; 5: >55 Percentage of time spent with other Brazilians (%) 50% or more of time spent with other Brazilians 1: Yes; 0: No. Feelings of discrimination. 1: Yes; 0: No. 0: Below high school; 1: High school; 2: College; 3: Graduate school. English proficiency prior to arriving in US 1: Yes; 0: No. Current English fluency I I 1: Good: 0: Poor. I ~ R A C O M I Desire to strengthen the Brazilian community I MONTHLY INCOME MALES RETURN US-CITIZEN I YRS-INUS I Years in the United States (in years) I I Figure 2. Variables used in models. - 1: Yes; 0: No. 1: <$1000; 2: $1000-$2000; 3: $2001-$3000; 4: $3001-$4000; 5: >$4000 1: Male; 0: Female. Desire to return to Brazil 1: Yes; 0: No. US citizenship? 1: Yes; 0: No. C I I Both logistic regression and linear regression models are used to analyze the data. Logistic regression is used when the dependent variable is binary (US-CITIZEN, RETURN, and ACT_WBRA50), while linear regression models are used when the dependent variable is continuous data (ACT-WBRA). Results The results from all four models (political, social and two psychological factors) assessing the causes of assimilation are reasonably good and are presented in Figure 3. In the first three models with US-CITIZEN, RETURN and ACT-WBRA50 respectively, as the dependent variables the X2 ranges from 27 to 69 percent, and the correct prediction rates range from 62 to 95 percent. In the last model where ACT-WBRA is used as the dependent variable, R2 and adjusted R2 are 28 percent and 16 percent respectively. For the logistic regressions (Models 1-3), values of EXP(B) greater than one indicates that the independent variable has a positive impact on the dependent variable, while a value of EXP(B) less than one indicates that the independent variable has a negative impact on the dependent variable. In contrast, in the linear regression model (Model 4) where ACT-WBRA is used as the dependent variable, the sign of the regression coefficient. indicates the direction of the relationship between the dependent variable and the independent variable: a positive sign for a positive relationship

Cox & Snell R* (%) 44 44 43 28.0' Nagelkerke R2 (%) 69.0 31.0 36.0 16.02 Corr. Red. rate 1 (%) 94.4 73.9 66.7 Corr. Pred. rate 2 (%) 62.5 71.4 72.7 Corr. Pred. rate 3 (%) 88.6 72.7 69.8 Figure 3. Results on the assimilation of Brazilian immigrants into the U.S. Notes: 1. In the model with ACT-WBRA, this represents the R2valuP. 2. In the model with ACT-WBRA, this represents the adjusted R2value. 3. Corr. Pred. Rate 1: Correct Prediction Rate for dependent variable=o; Corr. Pred. Rate 2: Correct Prediction Rate for dependent variable=l; Corr. Pred. Rate 3: Overall Prediction Rate. *: Significant at the level of 0.10. **: Significant at the level of 0.05. and a negative sign for a negative Brazilian immigrants (although this last relationship. one is counterintuitive to theories of In assessing the relationship assimilation). Indeed, MALES is the between gender and assimilation, it is only variable that is sipficant in all of clear in this study and consistent in all the models. - of the models that male immigrants are Regarding the relationship less likely to be assimilated than female between assimilation and other immigrants. Additionally, males are variables, the direction of the less likely to be U.S. citizens, less relationships concurs with the interested in returning to Brazil, and are segmented theory of assimilation and likely to spend less time with other supports the first hypothesis. In other

Dependant Variable: FEEL-BRACOM EXP(B) I Independent Variables I Model 1 I Model 2 I Model 3 I CONSTANT 1 104.98 1 67.704 1 48.3 I EDUCATION YRS-INUS MALES 0.176 1.259"" 0.494 0.161" 1.070 1.056 US-CITIZEN 1 0.022** RETURN 1 1.538-0.163" 1.083 1.135 Valid observations 44(27/ 17) 44(27/17) 43(26/17) Cox & Snell R2 (%) 33.0 18.4 18.9 Nagelkerke R2 (%) 44.8 25.0 25.5 Correct Pred. Rate 1 f%) 70.6 52.9 64.7 L Correct Pred. Rate 2 (%) 88.9 77.8 84.6 Correct Pred. Rate 3 (%) 81.8 68.2 76.7 Figure 4. Brazilian immigrants' assimilation arld their feeling about the local Brazilian community. ~otes: 1. Model 1 includes the independent variable US-CITIZEN as the indicator of assimilation; Model 2 includes the variable RETURN as the indicator of assimilation; and Model 3 use ACT-WRBA50 as the measurement of assimilation. 2. Correct Prediction Rate 1: Prediction Rate for dependent variable=o. Correct Prediction Rate 2: Prediction Rate for dependent variable=l. Correct Prediction Rate 3: Overall correct prediction rate. *: Variable is sigruficant at the level of 0.10. ": Variable is significant at the level of 0.05. words, those who have adopted many of the skills or attributes that facilitate assimilation are more Likely to be a U.S. citizen, less likely to want to return, and spend less time with other Brazilians. The independent variable AGE confirms that there is a positive relationship between age and U.S. citizenship and that, interestingly, AGE is also positively correlated with the desire/interest in returning to Brazil. More surprisingly perhaps, is that these sentiments are also true for individuals with a high level of education and English proficiency. These feelings are likely attributable to feelings of nostalgia and a longing for home, more so than any economic or political reasons which prompted the initial move. The models analyzing the relationship between immigrants' assimilation and their feeling about strengthening the local Brazilian community achieves moderate success with R2 ranging from 18.4 to 44.8 percent, and correct prediction rates from 52.9 to 88.9 percent (Figure 4). These results show that a person who is more assimilated to the mainstream society is less likely to expect/desire a stronger local Brazilian community, regardless if assimilation is measured by citizenship (US-CITIZEN), willingness to return Brazil (RETURN), or whether or not they spend 50 percent or more time with other Brazilians

(ACT-WE3RA50). Model 1, when US-CITIZEN is the key variable, shows a negative correlation between being a U.S. citizen and feelings about strengthening the Brazilian community. In Model 2 there is a positive relationship between the desire to retum to Brazil (RETURN) and feelings about strengthening the Brazilian community and the same is true in Model 3 where the binary independent variable ACT-WBRA50 is the key variable. The relationships between FEEL-BRACOM and the other independent variables measuring assimilation show mixed results. For instance, persons who have a higher level of education and speak better English and thus more likely to be assimilated are also less likely to expect or desire a stronger local Brazilian community. These results do not support the hypothesis that greater assimilation may lead to the desire to strengthen the Brazilian community. However, this same hypothesis is supported by Model 3, which shows that individuals who have been in the United States for a longer period tend to favor strengthening the local Brazilian community. There are several factors that may contribute to these mixed results. First, individuals with a high education and/or proficient English skills might have acquired these skills in Brazil and have not been in the United States for a very long time. Second, individuals might have been in the United States a long time but have not become proficient in English nor advanced their education and therefore desire a stronger Brazilian community. Third, those with strong English skills and a higher education are likely to have more associations outside of the Brazilian community and therefore have distanced themselves and have little desire to make it stronger. Finally, there are a few other results that are not consistent across the three models measuring feelings about strengthening the Brazilian community. The gender variable, MALES, shows different impacts in the three models. In Model 1 where US-CITIZEN is included, it has a negative impact on the dependent variable (FEEL-BRACOM). Male U.S. citizens are less likely to want the local Brazilian community strengthened. In Models 2 and 3 when, respectively, RETURN or ACT-WBRA50, are the key variables MALE is positively correlated; they would like to see the Brazilian community strengthened. This further reinforces the notion that those who have attained legal citizenship (by far the minority) do not wish the Brazilian community to be stronger and those who either want to retum to Brazil or spend the majority of their time with other Brazilians would like the community to be stronger. Discussion In Figure 1 we listed the variables used in the various models, however two of those variables 'Discrimination' and 'Income' were not used in any of the models but are nonetheless worthy of discussion. The respondents were asked whether they felt discriminated against in any way. Nearly all respondents said that they have not, nor have they ever been discriminated

against either socially or professionally. This is likely due to the large Latino population in South Florida where this type of discrimination is apparently absent or at least not detected as well as the ambition of many Brazilians to distance themselves from a community (i.e. identity), display a strong work ethic and otherwise avoid any stereotypes that may prompt discrimination. As can be expected, the income range among the Brazilians surveyed varied widely. However, it was surprising that many factors typically useful in predicting income did not hold true among this population. For example, among immigrant populations education, language proficiency, and citizenship will often give a reasonable indication of an individual's income. Among this population however, many people employed in manual labor positions (e.g. carpenter, house painter) that do not require advanced education nor, oftentimes, English proficiency were earning wages comparable to those working in professional occupations. The range of occupations among those surveyed includes: housekeepers, painters, graphic designers, delivery drivers, financial analyst, travel agent, and auto mechanic. One of the highest paid individuals in the survey is a travel agent who claims to earn approximately $8,000 per month yet is not proficient in English (this person is a U.S. citizen and does have a college education). There was one person in the survey with a graduate degree who worked as a babysitter and earned about $1,500 per month. This person spoke English fluently but was not a U.S. citizen. Twenty-nine of the 45 respondents have a college education yet the occupations among this sub-population include a babysitter, painter, realtor, travel agent, self-employed, small business owner, and accountant. The occupations of some of those with less than a college education include a tropical fish collector, maintenance worker, house cleaner, handyman, delivery driver, and landscaper. The salary range among this group, aside from a few outliers, is roughly comparable to those with a college degree. Only eight of the 45 respondents are U.S. citizens and this has very little bearing on income. Of the 45, eight responded that they are fluent in English, two of those are students and among the rest there is little difference in income compared with those who are less than fluent. In fact, those who claim to have poor English skills seem to earn incomes comparable to others included in the survey. Conclusion At the outset, this paper was primarily interested in the Brazilian immigration and assimilation experience and to build on earlier conceptualizations of the Brazilian immigrant community such as their "invisibility." As such, there are several important findings emanating from this research. First, segmented assimilation theory and its emphasis on exit conditions, entry context, and various human capital attributes seems to be a useful heuristic device for understanding the Brazilian assimilation

experience in South Florida. In this research, we have differentiated the immigrants through various means in an effort to better understand the individual characteristics that might foster greater assimilation. Our regression analyses on the assimilation process mostly confirmed expectations. Younger individuals with a relatively high level of education and who have some proficiency in English are likely to be more assimilated. One interesting finding was females tend to be better assimilated than males, even though they tend to spend a greater amount of their time with other Brazilians. This could be because many of the females in the study were relatively young and therefore likely to interact more with the host community and many females work in service sector occupations requiring greater social interaction. Those who have spent a greater number of years in the United States were likely to have stronger feelings about returning to Brazil. This can be attributed to some sense of nostalgia or homesickness as well as, perhaps, unfulfilled expectations in the country. Second, it seems that the "invisibility" of the Brazilian community in South Florida persists for a few factors attributable mostly to the immigrants themselves. Many Brazilians consciously choose to distance themselves from Brazilians (and Latinos in general) upon arrival. The majority of the respondents felt that this would lessen the likelihood of discrimination. Interestingly, our analysis showed that after the assimilation process has evolved, many hoped that the Brazilian community would become stronger. Third, Brazilians differ from many other ethnic groups where a large number of new immigrants earn a living serving their community through activities such as retail sales of "ethnic" items such as groceries or providing services targeted at their own community. These types of activities help to create a "visibility." This is not to say that these activities are absent from the Brazilian community, but they seem to be far less numerous and less geographically clustered than they are in other ethnic communities. Also, many Brazilians in this study work in occupations where the visibility tends to be low. With regard to feelings about strengthening the Brazilian community, those sentiments are weaker among those who are U.S. citizens and stronger among those who want to return to Brazil and those who already spend more than 50 percent of their time with other Brazilians. Research on assimilation and the immigrant experience in general, for good reason, is becoming increasingly complex. The experiences of the hundreds of immigrant groups that reside in the U.S. are highly diverse and warrant careful examination. This work almost necessarily begins with census data, but requires much more "on the ground" research to gain a more complete understanding of the myriad experiences that are shaping urban as well as rural areas. This modest study of Brazilian immigrants in South Florida is one step in that direction.

References Model, Suzanne and David Lapido. 1996. Alba, Richard D. and Victor Nee. 1997. Context and Opportunity: Minorities in London Rethinking Assimilation Theory for a New Era and New York. Social Forces 75(2): 485-510. of Immigration. Infernafional Migration Review 31(4): 826-874. Portes, Alejandro and Ruben G. Rumbaut. 1996. Boswell, Thomas D. 1993. The Cubanlmmigrant America: A Portrait (2nd ed,). Berkeley: University of California Press. American Homeland in Miami. Journal of Cultural Geography 13(2): 133-148. Portes, Alejandro and Alex Stepick. 1993. City on the Edge: The Transformation of Miami. Boswell, Thomas D. and James R. Curtis. 1984. Berkeley: University of California Press. ' The Cuban-American Experience. Totowa, NJ: Rowman and Allanheld. Portes, Alejandro and Min Zhou. 1993. The Brooke, James. 1993. In Brazil Wild Ways to Counter Inflation. New York Times, 25 July. de Souza, 0. 1996. Por Baixo do Pano. Veja, 10 April, pp. 60-62. Franko, Patrice. 1999. The Puzzle of Latin Amm'can Development. Lanham, MD: Rowman and Littlefield. New Second Generation: Segmented Assimilation and its Variants. Annals of the American Academy of Political and Social Science 530: 74-96. Rumbaut, Ruben G. 1997. Assimilation and its Discontents: Between Rhetoric and Reality. International Migration Review 31(4): 923-960. Sales, Teresa. 1998. Constructing an Ethnic Identity: Brazilian Immigrants in Boston, Mass. Migration World 26(5): 15-21. Goza, Franklin. 1994. Brazilian Immigration to North America. International Migration Review 28(1): 136-152. Stepick, Alex and Alejandro Portes. 1986. Flight Into Despair: A Profile of Recent Haitian Gans, Herbert J. 1997. Toward a Reconciliation of 'Assimilation' and 'Pluralism': The Interplay of Acculturation and Ethnic Retention. International Migration Review 31(4): 875-892. Lindstrom, David P. and Douglas S. Massey. 1994. Selective Emigration, Cohort Quality, and Models of Immigrant Assimilation. Social Science Research 23(4): 315-349. Logan, John R., Richard Alba, and Thomas McNulty. 1994. Ethnic Economies in Metropolitan Regions: Miami and Beyond. Social Forces 72(3): 691-724. Margolis, Maxine L. 1998. An Invisible Minority: Brazilians in New York City. Boston: Allyn and Bacon. Massey, Douglas S. 1985. Ethnic Residential Segregation: A Theoretical Synthesis and Empirical Review. Sociology and Social Research 69(3): 315-350. Refugees in South Florida. Migration Review 20(2): 329350, International Edward L. Jackiewicz and Yifei Sun are in the Department of Geography af California State University, Northridge. Their interests include ethnic geography and immigration. Lead author Edward Jackiewicz can be reached at <ed.jackiewicz@csun.edu>.

2003, Volume 5, Number 1-2 Geographical Voice Progress (?) in Geography David J. Wishart Articles Field Methods for Investigating Sense of Place Cary W. de Wit The Role of Place in the New American West: A Case Study of Colorado's Cuchara Valley Peter J. McCormick and Carolyn M. Daugherty The Becoming of the Land of the Underground Rain: Place as Paradox in Liberal, Kansas Patricia Solis An Assessment of Economic Change in the Northern Forest, 1971-1996 Kristopher D. White Alaska Identity Revisited Gregory Brown and Lilian Alessa Michigan's Changing Hispanic Population, 1990-2000 Evelyn D. Ravuri The Ties that Bind, or Not? The Assimilation of Brazilian Migrants in South Florida Edward L. Jackiewicz and Yifei Sun From the Field A Vision for the Future at Rio Rancho, New Mexico Hubert B. Stroud and Steve Tollefson Whiteclay, Nebraska: A Problem Landscape Matthew R. Engel A Linear Glance at Nacogdoches Douglas A. Hurt Special Section on Human Impacts on Wetlands, Christopher F. Meindl Guest Editor Geographical Voice Finding our way INTO Swamps 154 Christopher F. Meindl The Aboiteaux of Kamouraska: Tradition, Modernity, and Environmental Change in the Tidal Marshlands of Nineteenth and Early Twentieth Century Quebec 162 Matthew G. Hatvany Commissioners of Sewers and the Intensification of Agriculture in the Tantramar Marshlands of New Brunswick 183 Robert Summerby-Murray Transforming the Gardens of the Sea: The Manipulation of the Scarborough Marsh 205 Kimberly R. Sebold The All Too Human Geography of Wetlands in Early Nineteenth Century Delaware 218 David J. Grettler