Introduction
On Liberty Avenue in Richmond Hill, Queens, the storefronts shift
languages mid-block. Roti shops share walls with sari boutiques. Hindu
temples occupy converted rowhouses three doors down from mosques. The
signs are in English, but the names above them, the faces inside them,
and the music escaping through their doors belong to a world that traces
its origins not to South Asia directly but to the Caribbean coast of
South America. This is Little Guyana, the most concentrated
Indo-Guyanese community in the United States, and arguably in the world
outside Guyana itself. The community formed over two decades, between
the early 1970s and the close of the 1980s, through a migration stream
driven by political crisis in Guyana and pulled by the labor demand of a
restructuring New York economy. It consolidated in southern Queens,
specifically in Richmond Hill and South Ozone Park, in a pattern that
scholars have described, community members have documented, and
journalists have chronicled, but that no one has subjected to systematic
spatial analysis using individual-level census microdata. The geography
of that settlement remains unmeasured at the neighborhood scale. This
research provides the first systematic spatial analysis of its
concentration, how it changed over time, the economic mechanisms that
produced and sustained it, and how it compares to the theoretical models
that urban scholars use to understand immigrant spatial formations.
Statement of
Positionality
This study is a quantitative analysis of a community I come from.
Both of my parents’ families immigrated to New York City from Guyana in
the 1980s; they are, in the language of this study, Settler cohort
members. They arrived separately, fleeing the same political crossfire:
the PNC-PPP ethnic divide that structured Guyanese political life under
Burnham and made staying increasingly untenable for Indo-Guyanese
families caught between mandatory service, economic deterioration, and a
government that had foreclosed ordinary futures. They did not know each
other well in Guyana. They came to know each other in Queens. Both
families settled in multifamily homes in southern Queens, as tenants in
the same 2 to 4 family building stock that the Vertical Enclave Model
proposes. My parents met at Queens College. I am, in other words, a
downstream consequence of the processes this thesis documents. The
housing stock that the VEM describes as an affordability mechanism was
the housing stock my family lived in. The chain migration networks that
channeled successive Settler arrivals into the core zone are the
networks through which both families arrived. The community
infrastructure that made Queens legible as a destination for
Indo-Guyanese immigrants from the 1980s onward is the infrastructure I
grew up in. I am not an outside observer studying a community that
interests me.
I came to this project with cultural knowledge that no archive
provides, an intuitive understanding of what box hand is and how it
works, a felt sense of the weight of the PNC–PPP divide and what it
meant for ordinary families, and a recognition of Richmond Hill not as a
case study but as a place. This knowledge sharpened my intuitions
throughout and helped me ask questions of the data that an outside
researcher might not have considered. Insider status produces blind
spots as readily as it produces insight. I have had to work consciously
against the risk of confirming what I already believed, treating the
community narratives I absorbed growing up as established facts rather
than as hypotheses requiring empirical testing. The decision to ground
this study in census microdata rather than in ethnography was an
epistemological discipline: the data do not know what I think they
should show. They have pushed back on my assumptions, particularly
regarding the expected rates of self-employment, forcing a more honest
measurement of the community’s lived experience rather than a simple
narration of it.
I am also aware of the limits of what census data can hold. The
people I am counting and classifying were navigating a world that the
1990 PUMS does not record: the weight of leaving, the texture of
collective saving, the specific social trust through which a box hand
pool works, the particular mix of grief and pragmatism involved in
building a life in a new place after the old one had been made
impossible. My family lived in that world. This thesis can only gesture
toward it through the numbers that survived. I’ve been candid about that
gap, let the quantitative findings carry their weight, and acknowledged
where they run out. Finally, I am conscious that this project interprets
a community in terms of an ethnoburb, my proposed Vertical Enclave
Model, and spatial hardening, that the community itself did not use and
would not necessarily recognize. I have tried to use these frameworks as
analytical tools rather than as judgments, and to hold open the
possibility that the most important things about what my families and
their neighbors built in Richmond Hill are not the things that census
microdata is equipped to show. The community built something durable.
This research has tried to measure one dimension of how.
Research
Problem
The Indo-Guyanese settlement in Queens sits at the intersection of
three bodies of scholarship. The first is the community studies and
Caribbean diaspora literature, which offers rich ethnographic and
sociological accounts of identity, culture, and belonging in Richmond
Hill but does not systematically analyze the spatial formation or
evaluate it against competing theoretical frameworks (Bacchus, 2020;
Marinic, 2014; Roopnarine, 2018; Wainwright, 2012). The second is the
urban ethnic-geography literature, which has developed sophisticated
frameworks for understanding immigrant residential formations, including
the ethnic enclave, spatial assimilation, super-diversity, and ethnoburb
models, but has not applied any of them to the Indo-Guyanese Queens case
at the neighborhood scale (Li, 1998a, 1999, 2009; Massey & Denton,
1993; Vertovec, 2007; Wilson & Portes, 1980). The third is the
census-microdata literature on West Indian immigrants in New York, which
examines Caribbean-born residents at the borough level or through broad
identity categories rather than as distinct national-origin groups with
their own spatial formation dynamics (Crowder & Tedrow, 2001; Model,
2008; Waters, 1999).
Research
Questions
Three questions organize the empirical analysis. The first question
is spatial. How did the Indo-Guyanese community come to occupy a
spatially concentrated residential formation in Queens, New York City,
between 1970 and 1990, and how did that spatial distribution change over
time? This question asks where the community settled, how concentrated
that settlement was relative to the broader Queens population, and
whether the concentration grew, stabilized, or declined between the
first wave of arrivals in the 1970s and the second wave in the 1980s.
The second question is economic. What role did homeownership, and
specifically multi-family homeownership, play in anchoring the
residential formation in the core zone? This question asks whether the
distinctive 2 to 4 family housing stock of southern Queens, and the
capacity of owner-occupiers to receive rental income from co-resident
tenants, provided the economic mechanism through which pioneer settlers
achieved residential permanence and created the property-based
infrastructure into which subsequent arrivals were absorbed. The third
question concerns labor market integration. To what extent did
Indo-Guyanese residents in Queens participate in the broader
metropolitan labor market rather than in ethnic economic circuits
internal to the community? This question distinguishes between the
economic self-sufficiency predicted by enclave theory and the
metropolitan integration predicted by the ethnoburb framework, and asks
which pattern the census microdata supports.
Significance
This study provides the first systematic evidence that the
Indo-Guyanese settlement in Queens constitutes an ethnoburb in Li’s
(2009) sense, spatially concentrated, multiethnic in composition, and
integrated into the broader metropolitan economy, and identifies the
economic mechanism through which that formation was achieved and
sustained. The community is large, visible, and historically important
as one of the earliest and most enduring Caribbean immigrant settlements
in New York City. However, it has remained outside the quantitative
spatial analysis literature. This study provides the first systematic
measurement of the community’s residential concentration, temporal
dynamics, and housing economics. The second significance is theoretical.
The existing literature on urban ethnic spatial formation has primarily
developed from the experiences of East Asian, Latino, and European
immigrant communities in American metropolitan areas. The application of
these frameworks to an Anglophone Caribbean working-class community in a
dense inner-ring urban borough requires critical evaluation rather than
mechanical application. Where the existing frameworks do not transfer
directly, this study proposes an extension. The Vertical Enclave Model,
developed in Chapter 6, proposes that multi-family residential
homeownership can serve as the spatial anchoring mechanism for ethnoburb
formation in urban contexts where transnational commercial real estate
investment is absent. This extends Li’s ethnoburb framework to a class
of immigrant communities and urban contexts, filling gaps, not yet
addressed in existing literature. The third significance is contextual.
The Indo-Guyanese migration to Queens occurred against the backdrop of
one of the Caribbean’s most significant political crises. The
cooperative socialist experiment under Forbes Burnham produced economic
deterioration, ethnic marginalization, and a sustained outflow of
educated professionals and working-class families, reshaping the
demographic geography of southern Queens over two decades. Understanding
this migration and the spatial formation it produced contributes to the
historical record of Caribbean diasporic community formation in American
cities and to the broader scholarship on how political crises in small
developing states produce lasting urban geographies in metropolitan
receiving cities.
Theoretical
Framework
Four frameworks from urban ethnic geography organize this analysis.
Spatial assimilation theory predicts that residential concentration
declines as groups achieve socioeconomic mobility (Alba
& Nee, 1997). The ethnic enclave model attributes concentration
to economic self-sufficiency and high self-employment (Portes
& Manning, 1986). The super-diversity model suggests that high
ethnic fragmentation in neighborhoods like Queens prevents any single
group from achieving dominance (Vertovec,
2007). Li’s
(2009) ethnoburb predicts concentration paired with metropolitan
labor integration. This study adapts Li’s framework to the working-class
urban context of Queens. Unlike the capital-intensive ethnoburbs of the
Pacific Rim, the Indo-Guyanese formation relies on theVertical
Enclave Model (VEM). One plausible upstream mechanism for the VEM is
the rotating savings and credit association known as box hand in Guyana,
through which working-class households pooled capital for property
acquisition; this connection is developed and qualified in Section
8.7.
The Vertical
Enclave Model
The Vertical Enclave Model is an original theoretical framework
proposed by this study due gaps in the existing enclave literature. It
does not appear in the existing literature under this name or in this
formulation. The Vertical Enclave Model proposes that the spatial
anchoring mechanism for the Indo-Guyanese ethnoburb formation in Queens
was not a transnational commercial real estate investment, but rather
the strategic acquisition of multi-family residential properties in the
2 to 4 family housing stock of southern Queens by the Pioneer cohort of
settlers arriving in the 1970s. The 2 to 4 family building is the
dominant residential form in southern Queens, a product of the borough’s
early twentieth-century development as a streetcar suburb designed to
attract working-class and lower-middle-class owner-occupiers. These
buildings contain between two and four self-contained dwelling units,
typically with the owner occupying one unit and renting the remaining
units to tenants. The Census Bureau codes the monthly housing cost
variable OWNCOST as the total cost of ownership net of any rental income
received from co-resident tenants. An owner of a 2 to 4 family building
who rents out one or two units, therefore reports an OWNCOST that
reflects costs after offsetting rental income. The model proposes that
Pioneer settlers who acquired 2 to 4 family properties in Richmond Hill
and South Ozone Park were able to reduce their effective housing costs
substantially through rental income, enabling them to achieve
residential permanence in the core zone even on working-class incomes.
This permanence served two functions. First, it created a stable
residential anchor that attracted subsequent chain migrants into the
same geographic area, reinforcing spatial concentration through the
cumulative causation mechanism Massey (1990) describes. Second, it
created a supply of rental units within the core zone that Settler
arrivals could move into without competing in the broader Queens rental
market, lowering the cost and risk of initial settlement for the second
wave of migrants. The model is called vertical because the concentration
it describes is organized around the vertical stacking of households
within individual buildings rather than the horizontal clustering of
co-ethnic households in a commercial district. It is called an enclave
model not because it predicts economic self-sufficiency, but because the
residential structure it describes creates a form of housing-market
enclosure, a pool of co-ethnic rental supply within the core zone, that
operates through property ownership rather than through ethnic
entrepreneurship.
In contrast to commercial real estate investment, which necessitates
access to substantial business capital and transnational networks
possessed by Li’s initial ethnoburb settlers, the acquisition of
multi-family residential properties was accessible to working-class
households through standard mortgage financing. The owner-occupier
structure of these buildings, where the owner lives in one unit and
rents the remaining units to tenants, meant that the income stream
needed to service the mortgage was generated by the property itself
rather than by the owner’s wages alone. This is not simply a housing
affordability strategy. It is a wealth-building mechanism that creates a
pathway toward equity accumulation while simultaneously reducing the
effective cost of ownership below what the market would otherwise impose
on a working-class buyer. The VEM situates this mechanism within a
broader Caribbean tradition of informal economic solidarity. Hossein
(2017) documents the rotating savings and credit associations, known as
box hand in Guyana and sou-sou elsewhere in the Caribbean, through which
working-class Caribbean immigrant communities historically pooled
capital for large purchases, including property acquisition. Ardener
(1964) established the comparative framework within which these
practices are understood across cultures. The social infrastructure of
collective saving that Hossein documents provides the most plausible
account of how Pioneer settlers accumulated the down payments necessary
for multi-family acquisition on working-class incomes, a dimension of
the mechanism that census microdata cannot directly observe but that
secondary sources strongly suggest was operating in Richmond Hill during
the 1970s.
Data and
Methods Overview
This study uses a mixed-methods design, specifically the explanatory
sequential model described by Creswell and Plano Clark (2017), combining
quantitative analysis of census microdata with qualitative engagement
with secondary historical and community sources. The quantitative strand
uses IPUMS USA 5% Public Use Microdata Samples for 1980 and 1990 at the
PUMA level to analyze residential concentration, homeownership patterns,
housing economics, labor market integration, and cohort-based temporal
comparison for the Indo-Guyanese community in Queens. Sub-borough
spatial analysis is available for 1990 only, as the 1980 microdata
assigns all Queens respondents to a single County Group covering the
entire borough rather than to individual PUMAs. The temporal dimension
of spatial hardening is therefore addressed through borough-level
consolidation evidence and a within-1990 cohort comparison. IPUMS NHGIS
aggregate census data for 1980 and 1990 at the tract level supplements
the microdata analysis to document the demographic growth of the
foreign-born population in Queens across the study period.
Place-of-birth classifications in the STF3 tract-level files aggregate
small-country origins into broader regional categories; Guyana does not
appear as a distinct birthplace category at the tract level in either
the 1980 or 1990 files. Guyanese-born counts are therefore available
only through the IPUMS microdata at the PUMA level.
Chapter
Overview
Chapter 2 reviews literature on Indo-Guyanese migration, theories of
urban ethnic spatial formation, political history, brain drain, and
methods for measuring residential segregation for empirical analysis.
Chapter 3 outlines research design, data sources, sample, variables, and
methods. Chapter 4 covers historical and demographic context, political
causes of emigration, and demographic growth in Queens from 1970 to
1990. Chapter 5 documents the spatial hardening of the Richmond Hill and
South Ozone Park core zone over the study decade, establishing the 36.1
percentage-point gap in core zone concentration between Pioneer and
Settler cohorts that anchors the cumulative causation argument, with a
summary of spatial findings. Chapter 6 reports the results of the
Vertical Enclave Model regarding housing patterns and robustness tests,
with a summary of the housing economics findings. Chapter 7 reports data
on metropolitan integration, enclave rejection, employment, occupation,
and transportation, with a summary of labor market findings. Full
theoretical interpretation across all three results chapters is reserved
for Chapter 8. Chapter 8 interprets the empirical findings against all
four theoretical frameworks.
Chapter 8:
Discussion
8.1
Introduction
This chapter interprets the empirical findings reported in Chapters
5, 6, and 7 within the theoretical framework established in Chapter 2.
It addresses four questions. First, do the spatial concentration
patterns observed in Queens between 1980 and 1990 satisfy the criteria
for ethnoburb formation as defined by Li (1998, 2009)? Second, does the
this studies originally, proposed Vertical Enclave Model account for the
housing strategy that enabled Indo-Guyanese spatial consolidation?
Third, how do the labor market integration findings bear on the
distinction between the ethnoburb and the ethnic enclave? Fourth, what
do these findings contribute to the broader literature on Caribbean
immigrant settlement and urban ethnic spatial formation? The chapter
proceeds through these questions in order before addressing limitations
and directions for future research.
8.2 Competing
Frameworks Evaluated
Spatial assimilation predicts that residential concentration is a
transitional phenomenon declining as immigrant groups achieve
socioeconomic mobility and disperse into more integrated environments
(Alba & Nee, 1997; White, 1988). The data rejects this prediction.
Queens’ share of the citywide Indo-Guyanese total rose from 44.1 to 54.7
percent between 1980 and 1990, and within-Queens spatial concentration
hardened rather than dispersed across the decade: Settler cohort members
were concentrated in the core zone at 54.9 percent compared with 18.8
percent for Pioneers. Pioneer homeownership rates are substantially
higher than Settlers’ at 81.8 versus 64.3 percent, which is consistent
with the socioeconomic advancement spatial assimilation describes, but
that advancement did not produce departure from the ethnic settlement.
It produced peripheral zone dispersal within Queens, a pattern more
consistent with the place stratification framework Logan and Alba (1993)
propose, in which structural barriers and co-ethnic social
infrastructure constrain residential mobility even as incomes rise, than
with spatial assimilation’s prediction of outward dispersal into
integrated residential environments. The super-diversity framework
predicts extreme ethnic fragmentation in which no single group achieves
meaningful spatial concentration (Vertovec, 2007). Location Quotients of
3.29 for PUMA 5409 and 2.73 for PUMA 5412 directly contradict this
prediction. The core zone is multiethnic, Indo-Guyanese residents
constitute approximately one percent of its total population, but it is
not undifferentiated. Indo-Guyanese residents are present at two to
three times their expected Queens share in precisely the neighborhoods
that community scholarship identifies as Little Guyana. Super-diversity
may be an accurate characterization of Queens as a borough, where dozens
of national-origin communities coexist across a highly diverse
residential landscape, but as a prediction about the spatial behavior of
individual immigrant communities within that diversity, it does not fit
the evidence. The Indo-Guyanese case demonstrates that meaningful
co-ethnic spatial concentration can persist and intensify within a
super-diverse metropolitan borough, a finding consistent with Crul’s
(2016) argument that super-diversity and group-level concentration are
not mutually exclusive at different scales of analysis. A primary
limitation of using PUMS microdata is the geographic resolution; 1980
PUMAs cover broader areas than 1990 tracts, which may slightly obscure
the earliest “micro-clusters.”
8.3 Ethnoburb
Formation Confirmed
Li’s (1998, 2009) ethnoburb framework identifies three necessary
conditions for ethnoburb designation: spatial concentration of an ethnic
group within a suburban or outer-borough zone, multiethnic composition
of that zone rather than ethnic numerical dominance, and integration of
the group into the broader metropolitan economy rather than isolation
within an enclave labor market. The findings reported in Chapters 5 and
7 satisfy all three conditions for the Indo-Guyanese population of
Queens by 1990. The spatial concentration criterion is met. Location
Quotients of 3.29 for PUMA 5409 and 2.73 for PUMA 5412 indicate that
Indo-Guyanese residents were concentrated at levels two to three times
their expected share of the Queens population. The Index of
Dissimilarity of 0.409 places the Indo-Guyanese distribution in the
moderate concentration range, below the 0.60 threshold conventionally
associated with high segregation, and comparable to the Afro-Guyanese
index of 0.417. The moderate dissimilarity value requires careful
interpretation rather than dismissal. As documented in Sections 2.5 and
3.7, PUMA-level dissimilarity indices are structurally conservative
estimates of true neighborhood-level clustering. Wong (1997)
demonstrates that segregation indices calculated at coarser spatial
scales systematically understate the fine-grained clustering visible at
the tract or block-group level. The PUMA-level dissimilarity index of
0.409 should therefore be understood as a lower bound on the
concentration that tract-level analysis would reveal rather than as a
precise measure of neighborhood-level segregation. The LQ values of 2.73
and 3.29 confirm that concentration in the core zone is specific and
substantial even where borough-wide dissimilarity remains moderate:
Indo-Guyanese residents are present at two to three times their expected
share in precisely the two PUMAs that prior community scholarship
identifies as the heart of the Little Guyana settlement. A moderate
dissimilarity index at the PUMA level combined with high Location
Quotients in specific PUMAs is the expected signature of a community
that is concentrated in a defined sub-borough zone but not segregated
across the borough as a whole, which is precisely what the ethnoburb
framework predicts. The geographic contiguity of the two core PUMAs
reinforces this reading. South Ozone Park, Ozone Park, Richmond Hill,
and Woodhaven form a single connected sector of southern Queens anchored
by the Liberty Avenue commercial corridor and bounded by major arterials
and the Belt Parkway. The Indo-Guyanese community did not concentrate in
a single isolated pocket of a few city blocks, the spatial signature of
a classic ethnic enclave, but rather transformed a large,
multi-neighborhood sector of the outer borough. This spatial scale is a
defining feature of the ethnoburb as Li (2009) conceives it, and it is
directly visible in the Location Quotient gradient across Queens PUMAs:
concentration peaks in the contiguous core, falls to intermediate levels
in the immediately adjacent peripheral PUMAs, and approaches zero in the
more distant parts of the borough. The moderate dissimilarity value is
also not a weakness of the ethnoburb classification relative to the
enclave or segregation models. Li (2009) explicitly describes the
ethnoburb as a zone of relative concentration rather than involuntary
segregation. A dissimilarity index approaching 0.60 or above would be
more consistent with the involuntary residential segregation that Massey
and Denton (1993) document for Black Americans in hypersegregated
cities, or with the extreme spatial insularity of the classic ethnic
enclave, than with the voluntary and economically motivated
concentration the ethnoburb framework describes. The moderate
dissimilarity value is, in this sense, a confirmatory finding for the
ethnoburb classification. The multiethnic composition criterion is met.
Indo-Guyanese residents constituted a mean of 1.08 percent of the total
population across the two core PUMAs. Afro-Guyanese residents
constituted a mean of 2.37 percent. The remaining 96 to 97 percent of
core zone residents were neither Indo- nor Afro-Guyanese. The core zone
is a zone of concentration, not a zone of ethnic numerical dominance.
This is precisely the multiethnic urban landscape Li describes, in which
an immigrant group is economically and spatially consequential without
constituting a population majority. The duration-of-residence robustness
check reported in Section 5.8 and Appendix B bears on the interpretation
of the Pioneer cohort effect in ways that strengthen rather than
undermine the VEM argument. When years-in-US is added as a predictor
alongside Pioneer cohort membership, the Pioneer OR attenuates and
years-in-US emerges as the dominant predictor. This collinearity is
structural: Pioneer membership and years-in-US measure the same
underlying reality from different angles. The appropriate reading of
this result is that duration of residence is the mechanism through which
Pioneer cohort membership predicts homeownership, Pioneers had
accumulated more years in the US by 1990, enabling the capital
accumulation, credit establishment, and housing market navigation that
multi-family acquisition requires. The theoretical claim of the VEM is
thereby reframed from a claim about cohort identity to a claim about the
cumulative enabling effects of residential duration: it is time in the
US, concentrated among earlier-arriving Pioneers, that enabled the
housing strategy the VEM describes. This reframing is not a retreat from
the original argument. It is a more precise statement of the mechanism.
The spatial hardening reported in Chapter 5 adds a dynamic dimension to
the ethnoburb model. The 36.1 percentage point gap in core zone
concentration, 18.8 for Pioneers versus 54.9 percent for Settlers,
represents a longitudinal process of residential mobility rather than a
static cross-sectional difference. Rather than avoiding the core zone,
Pioneers likely used it as an initial entry point. Historical evidence
from Chapter 4, combined with high Pioneer homeownership (81.8%) and
arrival-era income data, suggests that Pioneers subsequently dispersed
to peripheral areas as their longer tenure enabled residential
upgrading. While Settlers reported a higher median household income
($69,600) than the Pioneer median ($49,824), this reflects nominal wage
growth across the decade rather than superior economic positioning. As
established in Chapter 6, Pioneer incomes at the time of arrival were
sufficient to support multi-family acquisitions. This confirms Massey’s
(1990) mechanism of cumulative causation: the Pioneer cohort established
the housing infrastructure and social networks that channeled Settlers
into the core, then moved outward as their economic positions matured,
while the Settler wave reinforced the concentration. Alternative
interpretations, specifically that Pioneers never concentrated in the
core, contradict established scholarship. Bacchus (2020), Marinic
(2014), and Arjoon (2000) all document the core zone as the primary
destination for Indo-Guyanese arrivals across both decades, facilitated
by specific information flows through print media. The evidence does not
support fundamentally different geographic orientations between the
cohorts; rather, it confirms a shared arrival pattern in which Pioneers
eventually achieved the residential mobility facilitated by property
ownership and longer tenure. The Queens consolidation at the borough
scale reinforces this reading. The Indo-Guyanese share of the NYC total
increased from 44.1 percent in 1980 to 54.7 percent in 1990, a gain of
10.6 percentage points, while the Bronx share declined from 26.1 to 21.5
percent and the Manhattan share collapsed from 6.65 to 0.56 percent.
Queens was not merely one of several destinations for Indo-Guyanese
immigrants. It became the dominant destination through the decade, with
Richmond Hill and South Ozone Park emerging as the primary zone of
settlement within Queens by the time the 1990 census was taken. The
ethnoburb hardened through successive migration rather than through
initial founding. This process is the empirical instantiation of the
cumulative causation mechanism Massey (1990) describes, operating at the
PUMA level: each successive migration wave, channeled by social networks
and the housing infrastructure established by earlier arrivals,
reinforced the spatial concentration of the core zone and intensified
the conditions that would attract the next wave in turn.
8.4 The Vertical
Enclave Model as Mechanism
The ethnoburb literature identifies the fact of spatial concentration
but has left underspecified the economic mechanism through which
immigrant households sustain ownership in high-cost metropolitan housing
markets. The Vertical Enclave Model proposed in for this study, fills
this gap for the Indo-Guyanese case. It posits that multi-family
residential properties are the structural unit through which the VEM
operates: owner-occupants offset high ownership costs through rental
income from co-resident tenants, enabling homeownership at income levels
that would not support single-family ownership. The Chapter 6 findings
provide strong empirical support for the VEM across five dimensions. The
evidentiary foundation of the VEM test is the Census Bureau’s OWNCOST
coding convention. The Census Bureau codes OWNCOST as the total monthly
cost of homeownership net of rental income received from co-resident
tenants (U.S. Census Bureau, 1993). This means that the OWNCOST
differential between multi-family and single-family owners documented in
Chapter 6 is not merely consistent with rental income offset as an
external inference, it is structurally produced by the rental income
that the Census Bureau coding convention captures directly in the
variable construction. The lower OWNCOST reported by multi-family owners
is not an artifact of lower property values, smaller mortgages, or
different financing arrangements. It is a direct reflection of the
rental income those owners received from tenant units within their
buildings. This coding convention transforms the OWNCOST differential
from a suggestive correlation into a structural measurement of the
mechanism the VEM describes. The ownership cost differential is large
and statistically robust. Median monthly ownership costs for
multi-family owners were $313 compared with $1,197 for single-family
owners, a differential of $884 (Wilcoxon W = 3,615, p < 2.2e-16). The
magnitude of this differential is not a marginal housing market
phenomenon. Multi-family owners face ownership costs that are
approximately one-quarter those of single-family owners at the median.
The differential is not income-driven. Within the bottom income
quintile, median ownership costs were $95 for multi-family owners and
$1,008 for single-family owners (Wilcoxon p < .001). Within the
middle income quintile, the differential was $516, with multi-family
owners at $283 and single-family owners at $799 (Wilcoxon p < .001).
The flatness of the multi-family OWNCOST profile across income
quintiles, ranging from only $95 in the bottom to $313 in the fourth
quintile, confirms that the differential is structural rather than a
consequence of multi-family owners occupying lower income positions.
Single-family costs rise with income. Multi-family costs do not. This
pattern is exactly what the VEM predicts when rental income offsets
mortgage and maintenance obligations regardless of the owner’s wage
income level. The cost-to-income burden is dramatically lower for
multi-family owners. Multi-family owners devote a median of 7.1 percent
of monthly household income to ownership costs. Single-family owners
devote 33.3 percent (Wilcoxon W = 3,391, p < 4.1e-15). The difference
of 26.2 percentage points represents a qualitatively different financial
relationship to housing rather than a marginal affordability advantage.
Multi-family owners are well within the conventional 30 percent
affordability threshold at the median. Single-family owners exceed it.
For a working-class immigrant household arriving in one of the most
expensive housing markets in the United States, the difference between a
7.1 percent and a 33.3 percent housing cost burden is the difference
between residential permanence and financial precarity. The rental
income linkage is consistent with the VEM interpretation and
structurally confirmed by the Census Bureau coding convention as
described above. Median gross rent among Indo-Guyanese renters in
multi-family buildings in the core zone was $521 per month (raw n = 11,
weighted N = 258) and $653 in the peripheral zone (raw n = 20, weighted
N = 327). A single rental unit covers approximately 50.8 percent of the
core zone OWNCOST differential of $1,026; rental income from two tenant
units in a 3-unit building would cover approximately 101.6 percent of
the differential. In the peripheral zone, a single rental unit at the
median RENTGRS of $653 covers approximately 94 percent of the $695
differential. The high peripheral coverage ratio, a single unit at the
median rent covering 94 percent of the differential, indicates that
peripheral multi-family owners were close to cost-neutral on their
housing even before accounting for a second tenant unit. The OWNCOST
differential between single- and multi-family properties is consistent
with the proposition that rental income from tenant units offsets a
substantial portion of ownership costs. The small cell sizes limit the
precision of these estimates, and they are reported accordingly as
supporting rather than primary evidence. The primary evidentiary basis
for the rental income interpretation remains the Census Bureau coding
convention, which structurally produces the differential rather than
merely being consistent with it. The robustness checks rule out the two
most plausible alternative interpretations. The household size
alternative holds that multi-family owners have larger households and
therefore face higher space needs met by the larger structure rather
than by income offset. The FAMSIZE Wilcoxon test rejects this:
multi-family owners have significantly smaller households at a median of
four compared with five for single-family owners (W = 5,971.5, p <
.001). The property value alternative holds that multi-family properties
are cheaper to own because they are worth less. The VALUEH Wilcoxon test
rejects this in terms of direction: multi-family properties show higher
median assessed values at $225,000 compared with $162,500 for
single-family properties (W = 2,758.5, p < .001). A measurement
caveat applies here. The 1990 census routed the home value question
differently across structure types: multi-family owner-occupants
reported the value of the entire structure, while single-family
owner-occupants reported the value of their unit alone. This routing
difference inflates reported multi-family values relative to
single-family values in ways that cannot be fully separated from true
price differences, and the VALUEH comparison must therefore be treated
as indicative rather than definitive. What the comparison does establish
is that the direction of the finding is inconsistent with the
cheaper-properties alternative explanation: multi-family owners were not
acquiring lower-valued properties. No causal weight is placed on the
magnitude of the VALUEH difference given the routing limitation.
8.5 The
Afro-Guyanese Parallel and the Generalizability of the VEM
The Afro-Guyanese parallel finding deserves sustained attention
beyond its function as a robustness check. Among Afro-Guyanese
homeowners in Queens in 1990, median OWNCOST was $316 for multi-family
owners and $1,215 for single-family owners, a differential of $899
(Wilcoxon p < 2.2e-16). The consistency of this pattern across both
Guyanese ethnic groups establishes that the VEM mechanism operates at
the level of housing stock economics rather than as a culturally
specific Indo-Guyanese strategy. Multi-family residential properties in
Queens in 1990 offered dramatically lower ownership costs relative to
single-family properties regardless of the ethnic identity of the owner.
The economics of the 2 to 4 family building, the rental income offset
structure that the Census Bureau OWNCOST coding convention captures,
were available to any household that acquired such a property,
irrespective of ethnicity, national origin, or cultural background. This
finding raises a question that the VEM must address: if the housing
stock economics were equally available to both Indo-Guyanese and
Afro-Guyanese households, why did the Indo-Guyanese community produce a
more spatially concentrated settlement pattern in the specific PUMAs of
Richmond Hill and South Ozone Park? The Afro-Guyanese dissimilarity
index of 0.417 is nearly identical to the Indo-Guyanese index of 0.409,
suggesting comparable levels of borough-wide distributional
concentration. But the community scholarship reviewed in Chapter 2 and
the historical evidence in Chapter 4 document Richmond Hill and South
Ozone Park as distinctively Indo-Guyanese in character in ways that the
aggregate dissimilarity measures do not fully capture. The answer to
this question lies not in the housing economics themselves but in the
social infrastructure through which access to those economics was
organized. Chain migration networks, the print media documented by
Arjoon (2000), the religious and cultural institutional infrastructure
documented by Marinic (2014), and the rotating credit associations
through which Pioneer settlers accumulated down payments, as discussed
in Chapter 1 and developed further in Section 8.7 below, all directed
Indo-Guyanese arrivals specifically to the core zone in ways that
translated the general availability of the VEM mechanism into a
spatially specific settlement pattern. The VEM is a necessary but not
sufficient condition for the Indo-Guyanese ethnoburb formation. The
social infrastructure of chain migration and ethnic community
organization is the condition that made the general mechanism spatially
specific.
8.6 Metropolitan
Labor Market Integration and the Enclave Rejection
The Chapter 7 findings close the loop on the ethnoburb versus enclave
distinction. An ethnic enclave, as theorized by Wilson and Portes (1980)
and elaborated by Portes and Bach (1985), is characterized by spatially
concentrated co-ethnic self-employment, ethnic ownership of enterprises
employing co-ethnic workers, and a labor market at least partially
insulated from the broader metropolitan economy. The Indo-Guyanese data
for Queens in 1990 are inconsistent with all three of these features.
Self-employment rates are 1.67 percent for Pioneers and 1.43 percent for
Settlers. Fisher’s exact tests return non-significant results for both
the cohort comparison (p = .646) and the zone comparison (p = 1.000).
With only 2 self-employed Pioneer respondents and 4 self-employed
Settler respondents in the raw sample, the absence of enclave
self-employment is not a matter of degree. It is categorical. The
Indo-Guyanese population in Richmond Hill and South Ozone Park in 1990
does not exhibit the labor market structure that the enclave model
requires. The occupational distribution confirms metropolitan
integration. Administrative/Clerical is the largest category for
Pioneers at 31.5 percent, with Service second at 16.9 percent and
Operators/Laborers third at 13.9 percent. Sales accounts for 12.3
percent of Pioneer employment and Professional/Technical for 4.9
percent. Together the three non-manual white-collar categories,
Administrative/Clerical, Sales, and Professional/Technical, account for
48.7 percent of Pioneer employment. Among Settlers,
Administrative/Clerical also leads at 22.5 percent, with Service at 20.1
percent, Operators/Laborers at 17.5 percent, and Sales at 13.5 percent,
with the same three non-manual categories accounting for 43.6 percent of
Settler employment. These are occupations distributed across the
metropolitan labor market, office, service, sales, and manual sectors,
not concentrated in co-ethnic enterprises within the residential zone.
Craft/Repair is modestly higher for Pioneers at 8.0 percent compared
with 6.1 percent for Settlers, and Farming/Forestry/Fishing accounts for
5.0 percent of Pioneer employment and 8.9 percent of Settler employment.
This distribution reflects the labor market profile of a working- and
lower-middle-class immigrant population incorporated into the
metropolitan economy at multiple occupational levels. The Indo-Guyanese
and Afro-Guyanese occupational distributions are statistically
distinguishable (X² = 266.68, df = 6, p < 2.2e-16). Afro-Guyanese
residents show higher Administrative/Clerical concentration (32.3
vs. 25.5 percent) and substantially higher Managerial/Executive shares
(9.3 vs. 5.4 percent), while Indo-Guyanese residents show higher Service
(18.5 vs. 14.3 percent) and Operators/Laborers (15.5 vs. 11.1 percent)
shares. This difference indicates that the two communities are
incorporated into different segments of the metropolitan labour market,
Afro-Guyanese more concentrated in the office and managerial tier,
Indo-Guyanese more concentrated in service and manual work, rather than
following an identical integration pathway. It does not, however,
constitute evidence of enclave employment for either group. Both
distributions reflect participation across the metropolitan economy, and
self-employment rates below two percent for both cohorts confirm the
categorical absence of enclave labour market structure. The enclave
model is rejected not by occupational similarity between the two groups
but by the absence of the co-ethnic self-employment concentration the
model requires. The commute mode data provide indirect spatial evidence
for metropolitan integration. Bus is the dominant mode for Pioneers at
58.0 percent, with Car/Truck/Van second at 26.2 percent. Settlers show a
more even split between Car/Truck/Van at 39.1 percent and Bus at 36.9
percent. No respondent in either cohort uses Subway/Rail. Bus and
private vehicle commuting to outer-borough and inner-suburban employment
destinations is consistent with the metropolitan integration thesis.
Workers commute out of the core zone to jobs distributed across the
metropolitan periphery rather than walking to co-ethnic enterprises
within the neighborhood. The Pioneer naturalization rate of 62.5 percent
compared with 24.8 percent for Settlers reflects duration of residence
rather than differential commitment to settlement. Pioneers arrived
before 1980 and had at minimum ten years in which to complete the
five-year residency requirement for naturalization. Most Settlers
arrived after 1980 and many had not yet accumulated sufficient residence
time by 1990. The high Pioneer naturalization rate is consistent with
the settlement permanence thesis and with the long-term residential
commitment that the ethnoburb framework predicts.
8.7 The VEM and
Caribbean Economic Culture
The Vertical Enclave Model as tested in Chapter 6 is an economic
model inferred from census microdata. It establishes that multi-family
ownership reduced median monthly ownership costs from $1,197 to $313 for
Indo-Guyanese homeowners in Queens, a differential of $884 that
translated into a housing cost burden of 7.1 percent of household income
rather than 33.3 percent. What the census data cannot establish is how
Pioneer settlers accumulated the down payments necessary to acquire
multi-family properties on working-class incomes in one of the most
expensive housing markets in the United States in the first place. This
question, the financing question that precedes the ownership question,
is where the quantitative evidence stops and the cultural context
begins. Hossein (2017) documents the rotating savings and credit
associations that have historically structured informal capital
accumulation in Caribbean immigrant communities. Known as box hand in
Guyana and sou-sou in Trinidad and other parts of the Caribbean, these
associations work as follows: a group of participants contributes a
fixed amount to a common pool on a regular schedule, weekly or monthly,
and each participant receives the full pool in rotation. The participant
who receives the pool in a given round gains access to a lump sum that
is a multiple of their individual contribution, precisely the capital
structure that a down payment requires. No interest is charged. No
formal credit history is needed. The mechanism runs on social trust and
community obligation rather than on the institutional infrastructure of
formal lending. Ardener (1964) established the comparative framework
within which these practices are understood across cultures, documenting
their presence across Africa, Asia, and the Caribbean as a
near-universal response to the exclusion of working-class populations
from formal credit markets. Light (1972) documents equivalent practices,
tanomoshi in Japanese American communities, hui in Chinese American
communities, in financing property acquisition and business formation in
the United States, situating the Caribbean ROSCA tradition within a
recognized pattern of immigrant wealth-building. The connection to the
VEM is not merely illustrative. Consider the arithmetic. A 2 to 4 family
property in southern Queens in the early 1970s, when Pioneer settlers
were establishing their foothold in Richmond Hill and South Ozone Park,
required a down payment that represented a substantial capital
mobilization challenge for households recently arrived from a
deteriorating economy. A box hand pool operating among ten participants
contributing $200 per month would generate a $2,000 lump sum per
rotation, sufficient, combined with individual savings, to begin
accumulating toward a down payment on the kind of property the VEM
describes. The social infrastructure of collective saving is, in this
sense, the upstream condition that enabled the downstream mechanism: box
hand generated the capital that purchased the multi-family property; the
multi-family property generated the rental income that offset the
mortgage; and the offset mortgage enabled the residential permanence
that anchored the ethnoburb. The three steps are analytically distinct
but causally linked, and Hossein’s documentation of these practices
specifically in Guyanese immigrant communities in New York closes the
geographic and cultural gap between the comparative literature and the
Queens case. It should be stated clearly, however, that the ROSCA
financing pathway remains inferred rather than demonstrated. The IPUMS
microdata establish what Pioneer settlers acquired; they say nothing
about how acquisition was financed. Hossein (2017) establishes that box
hand practices were prevalent in Guyanese New York communities in this
period, and Ardener (1964) and Light (1972) establish that equivalent
practices financed property acquisition in other immigrant communities,
but the connection between those documented practices and the specific
Pioneer down payments that enabled VEM adoption in Richmond Hill and
South Ozone Park is a plausible causal inference, not a proven pathway.
Oral history interviews with Pioneer cohort members, or archival records
from community lending institutions of the period, would be required to
confirm it. The quantitative and cultural evidence is consistent and
theoretically coherent; it does not rise to causal demonstration. This
dimension of the VEM cannot be directly tested with IPUMS microdata. The
census does not record participation in rotating credit associations.
But the convergence of three bodies of evidence, the $884 OWNCOST
differential that only makes sense if substantial rental income is being
received, the Hossein (2017) documentation of box hand practices in the
specific community this thesis studies, and the Light (1972) and Ardener
(1964) frameworks establishing that informal credit association
financing of property acquisition is a recognized pattern across
working-class immigrant groups, provides a strong circumstantial case
for the full causal chain the VEM describes. The quantitative evidence
establishes the outcome. The cultural literature establishes the
plausibility of the pathway. It should be acknowledged that confidence
in this circumstantial case is not drawn solely from those three bodies
of evidence, it is also informed by insider cultural knowledge of a
community in which rotating credit practices were a known and ordinary
feature of economic life. That knowledge does not substitute for direct
empirical evidence, but it does shape the prior with which the
convergence of indirect evidence is assessed, and intellectual honesty
requires naming it as such. Future research combining census analysis
with oral history, of the kind Bacchus (2020) conducts for the Richmond
Hill community more broadly, could test the connection directly. What
the present study can claim is that the two bodies of evidence are
consistent with one another in a way that is theoretically meaningful
and not coincidental.
8.8 Theoretical
Contributions
The Vertical
Enclave Model
The first and primary contribution is the Vertical Enclave Model as a
theoretical mechanism for ethnoburb consolidation. Prior work on
ethnoburbs has described the spatial outcome but left underspecified the
economic pathway through which immigrant households achieve and sustain
ownership in high-cost metropolitan housing markets. Logan et al. (2002)
note that the conditions enabling ethnic spatial concentration vary
substantially across immigrant groups and metropolitan contexts, and
that a descriptive framework alone cannot account for this variation.
The VEM addresses this limitation by identifying multi-family
residential property ownership with rental income offset as a specific,
empirically tractable mechanism. The $884 median OWNCOST differential,
the 7.1 versus 33.3 percent cost-to-income ratios, the structural
insensitivity of the multi-family cost profile across income quintiles,
and the congruence between ownership cost differentials and prevailing
rents together constitute a coherent economic structure that is testable
with standard census microdata available for any American metropolitan
area. The VEM is not specific to the Indo-Guyanese case. The
Afro-Guyanese replication confirms that the mechanism operates wherever
immigrant households have access to multi-family housing stock in a
competitive rental market. The outer boroughs of New York, the inner
suburbs of Los Angeles, the triple-deckers of Boston and Providence, and
the two-flats of Chicago all present housing stock configurations in
which the VEM logic may apply. Researchers studying other immigrant
communities in comparable housing markets can test whether the ownership
cost structure documented here replicates across different ethnic
groups, metropolitan areas, and time periods. The VEM is a hypothesis,
proposed by me, about the economics of multi-family ownership as an
affordability strategy, not a claim specific to Indo-Guyanese Queens. It
extends the ethnoburb framework from a descriptive account of spatial
outcomes into a mechanistic account of how those outcomes are produced,
filling the gap that critics of Li’s framework have identified.
The
Transit-Oriented Working-Class Ethnoburb
The third contribution is the transit-oriented working-class
ethnoburb as a distinct variant of Li’s original formation. Li’s
ethnoburb was suburban, automobile-dependent, and populated by a
professional and capitalized immigrant class investing transnational
Pacific Rim capital in commercial real estate. The Indo-Guyanese Queens
settlement differs from Li’s original in nearly every structural
respect: it is urban rather than suburban, transit-oriented rather than
automobile-dependent, working-class rather than professional, Caribbean
rather than Pacific Rim, and anchored by residential rather than
commercial property investment. That ethnoburb spatial outcomes emerged
under these conditions suggests the framework describes a more general
logic of immigrant spatial formation than its origins imply. The
transit-oriented working-class ethnoburb is proposed as a distinct
variant of the formation warranting recognition in its own right rather
than as a modified application of Li’s original. The necessary and
sufficient conditions for this variant can be stated as follows. A
transit-oriented working-class ethnoburb requires: a dense inner-ring
urban morphology with attached or semi-detached multi-unit residential
stock; a public transit network providing metropolitan labor market
access without automobile ownership; a working-class or
lower-middle-class immigrant population with access to the VEM mechanism
through multi-family property acquisition; and a chain migration network
capable of channeling successive arrival waves into the established core
zone. When these conditions are met, the ethnoburb formation process can
operate without transnational commercial capital, without suburban
spatial morphology, and without the professional class composition of
Li’s original San Gabriel Valley case. The Queens settlement is the
first documented instance of this variant, but the housing stock and
transit conditions it requires are present in many American inner-ring
urban neighborhoods that have received working-class immigrant
populations in the post-1965 period.
The Caribbean
Ethnoburb
The fourth contribution is the application of the ethnoburb framework
to a Caribbean immigrant community of South Asian descent. Li’s original
ethnoburb research focused on Chinese Americans in the San Gabriel
Valley. The Indo-Guyanese Queens case demonstrates that the ethnoburb
formation process is not specific to a particular ethnic or racial
origin or to Pacific Rim transnational networks. Post-1965 immigrant
communities organized by chain migration, displaced by political
instability, and positioned in the outer boroughs of a high-cost
metropolitan area can produce ethnoburb spatial outcomes through the VEM
mechanism, provided they have access to the multi-family housing stock
that enables the rental income offset strategy. This extends the
empirical base of the ethnoburb concept across two dimensions
simultaneously: ethnic and racial origin, and class and capital
endowment. The Indo-Guyanese case is working-class where Li’s ethnoburb
was professional, Caribbean where Li’s ethnoburb was Pacific Rim, and
urban where Li’s ethnoburb was suburban. The ethnoburb outcome emerged
nonetheless. The framework is more general than its origins imply, and
the Indo-Guyanese Queens settlement is the evidence for that generality.
#8.9 Limitations and Future Research Four limitations bound the
conclusions of this thesis. The first is the 1980 PUMA geography
constraint. PUMA-level spatial disaggregation is available only for
1990. The within-Queens spatial distribution of the 1980 Indo-Guyanese
population cannot be recovered from the IPUMS microdata. The spatial
hardening argument therefore rests on within 1990 cohort comparison and
borough-level consolidation evidence rather than on direct sub-borough
geographic tracking across census years. Future research using NHGIS
block group or tract data for 1980 could extend the spatial analysis to
the full decade and provide a direct rather than indirect measure of
spatial hardening. The second is the reliance on OWNCOST as the primary
measure of the VEM mechanism. Although the Census Bureau’s OWNCOST
coding convention confirms that the observed differential is
structurally produced by rental income received rather than merely
consistent with it as an external inference, INCTOT in the IPUMS extract
does not itemize rental income separately from other income sources. A
direct test of the rental income magnitude would require rental income
data at the household level, available in later American Community
Survey extracts but not in the 1990 PUMS. The RENTGRS linkage test
provides the closest available approximation and rests on a core zone
renter cell of 11 raw cases (weighted N = 258), a cell size that limits
the precision of the estimate. The Census Bureau OWNCOST coding
convention reduces but does not eliminate the dependence on this cell.
The third is the small sample size. The raw Indo-Guyanese Queens sample
of 295 cases in 1990 is sufficient for the weighted analyses reported
here but limits the power of subgroup analyses. Replication with a
larger sample, if available from alternative sources, would strengthen
confidence in the findings. Future research using the American Community
Survey, which provides larger annual samples and finer geographic detail
than the decennial PUMS, could extend the analysis beyond 1990 and
provide a more precise spatial and temporal picture of the settlement’s
evolution. The fourth is the absence of qualitative data. The VEM is an
economic model inferred from census microdata. Whether Indo-Guyanese
homeowners in Richmond Hill and South Ozone Park understood their
housing strategy in the terms the model describes, whether multi-family
ownership was a deliberate affordability mechanism or an artifact of
Queens housing stock availability, and how community networks
facilitated property acquisition are questions that microdata cannot
answer. Future research combining census analysis with oral history and
ethnographic observation would provide a richer account of the
mechanisms identified here. The convergence of Hossein’s (2017)
documentation of rotating credit associations in Guyanese immigrant
communities and Ardener’s (1964) comparative framework suggests that the
social infrastructure for the VEM strategy may have deeper roots in
Caribbean economic culture than the census data alone can establish, and
that qualitative investigation of this connection would be a productive
direction for future research.
LAYER 3:
SAVE ENVIRONMENT AND VERIFY
save(
# Data objects
indo_queens, indo_guyanese, afro_guyanese, afro_queens_1990,
cohort_analysis, owners_1990, owners_clean, indo_owners_q, afro_owners_q,
lq_table, queens_total_1990, indo_by_puma, afro_by_puma_1990,
# Results objects
cascade, temporal_hardening, cohort_zone, cohort_summary_fixed,
arrival_pulse, multiethnic_table,
reg_sample, model_1, model_1_or,
mf_reg_sample, model_mf_a, model_mf_b, model_mf_c, table_6_4,
occ_final, occ_comparison_final,
tranwork_cohort, tranwork_zone, citizen_cohort,
owncost_diff_wide, diff_combined, rentgrs_zone,
fisher_se_cohort, fisher_se_zone,
file = "guyana_queens_analysis.RData"
)
cat("\nEnvironment saved to guyana_queens_analysis.RData\n")
##
## Environment saved to guyana_queens_analysis.RData
Final
verification
cat("1. OWNCOST differentials:\n")
## 1. OWNCOST differentials:
print(owncost_diff_wide %>% select(Core_Area, Multi, Single, Differential))
## # A tibble: 2 × 4
## Core_Area Multi Single Differential
## <chr> <dbl> <dbl> <dbl>
## 1 Core 324 1350 1026
## 2 Peripheral 313 1008 695
cat("\n2. Figure 6.1 row count (expected 8 — Q1-Q4 × 2 structure types):",
nrow(owncost_all_q), "\n")
##
## 2. Figure 6.1 row count (expected 8 — Q1-Q4 × 2 structure types): 8
cat(" Table 6.1 row count (expected 8 — Q1-Q4 × 2 structure types):",
nrow(table_6_1), "\n")
## Table 6.1 row count (expected 8 — Q1-Q4 × 2 structure types): 8
stopifnot("Table 6.1 row count unexpected — check quintile filter" =
nrow(table_6_1) == 8)
owners_clean <- owners_1990 %>%
# 1. Ensure Household_Income is calculated at the household level
group_by(SERIAL) %>%
mutate(Household_Income = sum(INCTOT, na.rm = TRUE)) %>%
ungroup() %>%
# 2. Apply the filters to reach the N=131 analytical sample
filter(
Core_Area == "Core", # Only include the Richmond Hill/S.Ozone Park core
Household_Income < 9999999, # Exclude top-coded income
OWNCOST < 99999, # Exclude top-coded costs
OWNCOST > 0 # Exclude cases with no cost data
) %>%
# 3. Re-calculate quintiles based ONLY on this specific 131-case sample
mutate(
Income_Quintile = ntile(Household_Income, 5)
)
# 4. Verify the counts
cat("Current owners_clean N:", nrow(owners_clean), "\n")
## Current owners_clean N: 38
print(table(owners_clean$Is_Multi_Family))
##
## FALSE TRUE
## 17 21
cat("\n3. Multi-family rates, owners only",
"(expected Pioneer ~29%, Settler ~36%):\n")
##
## 3. Multi-family rates, owners only (expected Pioneer ~29%, Settler ~36%):
print(cohort_summary_fixed %>% select(Cohort, multi_family_rate_owners))
## # A tibble: 2 × 2
## Cohort multi_family_rate_owners
## <chr> <dbl>
## 1 Pioneer 29.0
## 2 Settler 35.9
cat("\n4. Fisher p-values:\n")
##
## 4. Fisher p-values:
cat(" Self-employment by cohort:", round(fisher_se_cohort$p.value, 3), "\n")
## Self-employment by cohort: 0.646
cat(" Self-employment by zone: ", round(fisher_se_zone$p.value, 3), "\n")
## Self-employment by zone: 1
cat("\n5. Pioneer homeownership OR and 95% CI:\n")
##
## 5. Pioneer homeownership OR and 95% CI:
print(model_1_or %>% filter(Predictor == "Pioneer Cohort") %>%
select(Odds_Ratio, CI_Lower, CI_Upper, P_Value))
## Odds_Ratio CI_Lower CI_Upper P_Value
## Is_Pioneer 2.784257 2.415321 3.217935 1.692427e-44
cat("\n6. Table 6.4 — Settler OR attenuation (Model A → B when Core_Area added):\n")
##
## 6. Table 6.4 — Settler OR attenuation (Model A → B when Core_Area added):
cat(" Model A Settler OR:", round(exp(coef(model_mf_a)["Is_Settler"]), 3), "\n")
## Model A Settler OR: 1.369
cat(" Model B Settler OR:", round(exp(coef(model_mf_b)["Is_Settler"]), 3), "\n")
## Model B Settler OR: 1.03
cat(" Core Zone OR (Model B):", round(exp(coef(model_mf_b)["Is_Core"]), 3), "\n")
## Core Zone OR (Model B): 2.344
cat(" Core Zone OR (Model C):", round(exp(coef(model_mf_c)["Is_Core"]), 3), "\n")
## Core Zone OR (Model C): 2.644
cat("\n7. Core zone weighted N (expected Core ~3,178, Peripheral ~3,670):\n")
##
## 7. Core zone weighted N (expected Core ~3,178, Peripheral ~3,670):
indo_queens %>% ungroup() %>% select(-any_of("by")) %>%
filter(YEAR == 1990) %>%
group_by(Core_Area) %>%
summarise(weighted_n = sum(PERWT), .groups = "drop") %>% print()
## # A tibble: 2 × 2
## Core_Area weighted_n
## <chr> <dbl>
## 1 Core 3178
## 2 Peripheral 3670
cat("\n8. OCC table file check:\n")
##
## 8. OCC table file check:
cat(" tables/table_7_1_occ_cohort.csv exists (should be TRUE):",
file.exists("tables/table_7_1_occ_cohort.csv"), "\n")
## tables/table_7_1_occ_cohort.csv exists (should be TRUE): TRUE
cat(" tables/table_7_2_occ_comparison.csv exists (should be TRUE):",
file.exists("tables/table_7_2_occ_comparison.csv"), "\n")
## tables/table_7_2_occ_comparison.csv exists (should be TRUE): TRUE
cat(" tables/table_6_4_mf_logistic_nested.csv exists (should be TRUE):",
file.exists("tables/table_6_4_mf_logistic_nested.csv"), "\n")
## tables/table_6_4_mf_logistic_nested.csv exists (should be TRUE): TRUE
cat(" tables/table_5_1_lq_table.csv exists (should be TRUE):",
file.exists("tables/table_5_1_lq_table.csv"), "\n")
## tables/table_5_1_lq_table.csv exists (should be TRUE): TRUE
for (f in c("tables/appendix_d_occ_cohort.csv",
"tables/appendix_d_occ_comparison.csv",
"tables/appendix_d_occ_final.csv",
"tables/appendix_d_occ_comparison_final.csv")) {
cat(" ", basename(f), "exists (should be FALSE):", file.exists(f), "\n")
}
## appendix_d_occ_cohort.csv exists (should be FALSE): FALSE
## appendix_d_occ_comparison.csv exists (should be FALSE): FALSE
## appendix_d_occ_final.csv exists (should be FALSE): FALSE
## appendix_d_occ_comparison_final.csv exists (should be FALSE): FALSE
cat("\n9. Figure file check — definitive thesis numbering:\n")
##
## 9. Figure file check — definitive thesis numbering:
thesis_figures <- c(
"figures/figure_4_1_arrival_pulse.png",
"figures/figure_5_1_lq_bar.png",
"figures/figure_5_2_borough_distribution.png",
"figures/figure_5_3_queens_share_trend.png",
"figures/figure_5_4_homeownership_trend.png",
"figures/figure_5_5_cohort_zone.png",
"figures/figure_5_6_core_zone_cohort_comparison.png",
"figures/figure_5_7_logistic_forest.png",
"figures/figure_6_1_mf_rate_cohort_zone_group.png",
"figures/figure_6_2_owncost_quintile.png",
"figures/figure_6_3_rentgrs_linkage.png",
"figures/figure_6_4_owncost_differential_by_group.png",
"figures/figure_7_1_occ_cohort.png",
"figures/figure_7_2_occ_diverging.png",
"figures/figure_7_3_tranwork_cohort.png",
"figures/figure_7_4_tranwork_group_comparison.png",
"figures/figure_7_5_citizenship_cohort.png"
)
for (f in thesis_figures) {
cat(" ", basename(f), "—", ifelse(file.exists(f), "OK", "MISSING"), "\n")
}
## figure_4_1_arrival_pulse.png — OK
## figure_5_1_lq_bar.png — OK
## figure_5_2_borough_distribution.png — OK
## figure_5_3_queens_share_trend.png — OK
## figure_5_4_homeownership_trend.png — OK
## figure_5_5_cohort_zone.png — OK
## figure_5_6_core_zone_cohort_comparison.png — OK
## figure_5_7_logistic_forest.png — OK
## figure_6_1_mf_rate_cohort_zone_group.png — OK
## figure_6_2_owncost_quintile.png — OK
## figure_6_3_rentgrs_linkage.png — OK
## figure_6_4_owncost_differential_by_group.png — OK
## figure_7_1_occ_cohort.png — OK
## figure_7_2_occ_diverging.png — OK
## figure_7_3_tranwork_cohort.png — OK
## figure_7_4_tranwork_group_comparison.png — OK
## figure_7_5_citizenship_cohort.png — OK
References
Alba, R. D., & Nee, V. (1997). Rethinking assimilation theory for
a new era of immigration. International Migration Review, 31(4),
826–874. https://doi.org/10.2307/2547416
Allen, J. P., & Turner, E. (1996). Spatial patterns of immigrant
assimilation. The Professional Geographer, 48(2), 140–155. https://doi.org/10.1111/0033-0124.00016
Ardener, S. (1964). The comparative study of rotating credit
associations. Journal of the Royal Anthropological Institute, 94(2),
201–229. https://doi.org/10.2307/2844382
Arjoon, R. (2000).The Indo-Guyanese diaspora: Evolution of a
community. [Unpublished master’s thesis]. City University of New
York.
Bacchus, N. S. (2020). Belonging and boundaries in Little Guyana:
Conflict, culture, and identity in Richmond Hill, New York. Ethnicities,
20(5), 896–914. https://doi.org/10.1177/1468796820930721
Bartels, D. (1980). Ethnicity, ideology, and class struggle in
Guyanese society. Anthropologica, 22(1), 45–60. https://doi.org/10.2307/25605038
Burnham, F. (1970). A destiny to mould: Selected speeches by the
Prime Minister of Guyana (C. A. Nascimento & R. A. Burrowes,
Comps.). Africana Publishing Corporation.
Cambridge, V. C. (2015). The 1970s: Making the small man a real man.
In Musical life in Guyana (pp. 181–208). University Press of
Mississippi. https://doi.org/10.14325/mississippi/9781628460117.003.0008
Conway, D. (1990). Migration in the Caribbean. In R. B. Potter (Ed.),
Urbanisation, planning and development in the Caribbean (pp. 174–196).
Mansell Publishing.
Creswell, J. W., & Plano Clark, V. L. (2017). Designing and
conducting mixed methods research (3rd ed.). SAGE Publications.
Crowder, K. D., & South, S. J. (2005). Race, class, and changing
patterns of migration between poor and nonpoor neighborhoods. American
Journal of Sociology, 110(6), 1715–1763. https://doi.org/10.1086/428686
Crowder, K. D., & Tedrow, L. M. (2001). West Indians and the
residential landscape of New York. In N. Foner (Ed.), Islands in the
city: West Indian migration to New York (pp. 81–114). University of
California Press.
Crul, M. (2016). Super-diversity vs. assimilation: How complex
diversity in majority-minority cities challenges the assumptions of
assimilation. Journal of Ethnic and Migration Studies, 42(1), 54–68. https://doi.org/10.1080/1369183X.2015.1061425
Curless, G. (2023). Co-operative citizens? Development, work and
protest in Guyana, c. 1970–1985. International Review of Social History,
68(3), 389–428. https://doi.org/10.1017/S0020859023000603
Duncan, O. D., & Duncan, B. (1955). A methodological analysis of
segregation indexes. American Sociological Review, 20(2), 210–217. https://doi.org/10.2307/2088328
Eldridge, D. (1983). The impact of brain drain on development: A case
study of Guyana [Review of the book The impact of brain drain on
development: A case study of Guyana, by M. J. Boodhoo & A. Baksh].
Public Administration and Development, 3(1), 82–83. https://doi.org/10.1002/pad.4230030119
Fein, D. J. (1990). Racial and ethnic differences in U.S. census
omission rates. Demography, 27(2), 285–302. https://doi.org/10.2307/2061463
Foner, N. (Ed.). (2001). Islands in the city: West Indian migration
to New York. University of California Press.
Glasgow, R. A. (1970). Guyana: Race and politics among Africans and
East Indians. M. Nijhoff.
Glaser, W. A., & Habers, G. C. (1978). The brain drain:
Emigration and return. Findings of a UNITAR multinational comparative
survey of professional personnel of developing countries who study
abroad. Pergamon Press.
Gopaul, N. K. (1997). Resistance and change: The struggles of
Guyanese workers. Ian Randle Publishers.
Hintzen, P. C. (1989). The costs of regime survival: Racial
mobilization, elite domination and control of the state in Guyana and
Trinidad. Cambridge University Press.
Hope, K. R. (1972). The role of government expenditure in the
economic development in Guyana (1960–1970). The American Economist,
16(2), 166–174. https://doi.org/10.1177/056943457201600225
Hossein, C. S. (2017). Building economic solidarity: Caribbean ROSCAs
in Jamaica, Guyana, and Haiti. In C. S. Hossein (Ed.), The Black social
economy in the Americas (pp. 79–95). Palgrave Macmillan. https://doi.org/10.1057/978-1-137-60047-9_5
Jackson, S. N. (2012). From myth to market: Burnham’s co-operative
republic. In Creole indigeneity (pp. 111–140). University of Minnesota
Press. https://doi.org/10.5749/minnesota/9780816677757.003.0005
Jagan, C. (1980). The West on trial: The fight for Guyana’s freedom
(Rev. ed.). Seven Seas Publishers.
Khemraj, T. (2015). The colonial origins of Guyana’s
underdevelopment. Social and Economic Studies, 64(3/4), 151–185.
Lee, E. S. (1966). A theory of migration. Demography, 3(1), 47–57. https://doi.org/10.2307/2060063
Li, W. (1998a). Anatomy of a new ethnic settlement: The Chinese
ethnoburb in Los Angeles. Urban Studies, 35(3), 479–501. https://doi.org/10.1080/0042098984871
Li, W. (1998b). Ethnoburb versus Chinatown: Two types of urban ethnic
communities in Los Angeles. Cybergeo: European Journal of Geography. https://doi.org/10.4000/cybergeo.1018
Li, W. (1998c). Los Angeles’s Chinese ethnoburb: From ethnic service
center to global economy outpost. Urban Geography, 19(6), 502–517. https://doi.org/10.2747/0272-3638.19.6.502
Li, W. (1999). Building ethnoburbia: The emergence and manifestation
of the Chinese ethnoburb in Los Angeles’ San Gabriel Valley. Journal of
Asian American Studies, 2(1), 1–28. https://doi.org/10.1353/jaas.1999.0009
Li, W. (Ed.). (2006). From urban enclave to ethnic suburb: New Asian
communities in Pacific Rim countries. University of Hawaiʻi Press.
Li, W. (2008). Ethnoburb. In W. Li (Ed.), Ethnoburb: The new ethnic
community in urban America (pp. 1–20). University of Hawaiʻi Press. https://doi.org/10.21313/hawaii/9780824830656.003.0002
Li, W. (2009). Ethnoburb: The new ethnic community in urban America.
University of Hawaiʻi Press.
Light, I. (1972). Ethnic enterprise in America: Business and welfare
among Chinese, Japanese, and Blacks. University of California Press.
Light, I., & Karageorgis, S. (1994). Economic saturation and
immigrant entrepreneurship. In L. Isralowitz & I. Light (Eds.),
Immigration and absorption: Issues in a multicultural perspective
(pp. 89–108). Ben-Gurion University of the Negev.
Light, I., & Rosenstein, C. (1995). Race, ethnicity, and
entrepreneurship in urban America. Aldine de Gruyter.
Light, I., Bhachu, P., & Karageorgis, S. (1993).Migration
networks and immigrant entrepreneurship. In I. Light & P. Bhachu
(Eds.), Immigration and entrepreneurship: Culture, capital, and networks
(pp. 25–49). Transaction Publishers.
Logan, J. R., & Alba, R. D. (1993). Locational returns to human
capital: Minority access to suburban community resources. Demography,
30(2), 243–268. https://doi.org/10.2307/2061895
Logan, J. R., Zhang, C., & Alba, R. D. (2002).Immigrant enclaves
and ethnic communities in New York and Los Angeles. American
Sociological Review, 67(2), 299–322. https://doi.org/10.2307/3088897
Marinic, G. (2014). Domestic deities: Indo-Caribbean spatial
territorialization and sacred space in South Richmond Hill, Queens.
Material Religion, 10(1), 28–55. https://doi.org/10.2752/175183414X13834769373168
Mars, P. (1998). Ideology and change: The transformation of the
Caribbean Left. University of the West Indies Press.
Massey, D. S. (1990). Social structure, household strategies, and the
cumulative causation of migration. Population Index, 56(1), 3–26. https://doi.org/10.2307/3644186
Massey, D. S., & Denton, N. A. (1988). The dimensions of
residential segregation. Social Forces, 67(2), 281–315. https://doi.org/10.2307/2579183
Massey, D. S., & Denton, N. A. (1993). American apartheid:
Segregation and the making of the underclass. Harvard University
Press.
Minnesota Population Center. (2023a). CONSPUMA: Consistent public use
microdata area. IPUMS USA. https://usa.ipums.org/usa-action/variables/CONSPUMA
Minnesota Population Center. (2023b). ipumsr: An R interface for
IPUMS data (R package version 0.7.2). University of Minnesota. https://cran.r-project.org/package=ipumsr
Miyares, I. M. (2004). From exclusionary covenant to ethnic
hyperdiversity in Jackson Heights, Queens. Geographical Review, 94(4),
462–483. https://doi.org/10.1111/j.1931-0846.2004.tb00178.x
Model, S. (2008). West Indian immigrants: A Black success story?
Russell Sage Foundation.
Modeste, N. C. (2021). Efficiency-adjusted public capital and
economic growth in Guyana: A cointegration analysis. Atlantic Economic
Journal, 49(2), 187–199. https://doi.org/10.1007/s11293-021-09714-5
Mohabir, N., & Cummings, R. (2019). An archive of loose leaves:
An interview with Frank Birbalsingh. Small Axe: A Journal of Criticism,
23(3), 104–118. https://doi.org/10.1215/07990537-7912358
Niland, J. R. (1970). The Asian engineering brain drain: A study of
international relocation into the United States from India, China,
Korea, Thailand and Japan. Heath Lexington Books.
NYC Department of City Planning. (n.d.). Residence districts: R4
zoning. New York City Government. https://www.nyc.gov/site/planning/zoning/districts-tools/r4-r4-1-r4a-r4b.page
Pebesma, E. (2018). Simple features for R: Standardized support for
spatial vector data. The R Journal, 10(1), 439–446. https://doi.org/10.32614/RJ-2018-009
Pedersen, T. L. (2024). patchwork: The composer of plots (R package
version 1.2.0). https://cran.r-project.org/package=patchwork
Portes, A., & Bach, R. L. (1985). Latin journey: Cuban and
Mexican immigrants in the United States. University of California
Press.
Portes, A., & Jensen, L. (1987). What’s an ethnic enclave? The
case for conceptual clarity. American Sociological Review, 52(6),
768–771. https://doi.org/10.2307/2095551
Portes, A., & Jensen, L. (1989). The enclave and the entrants:
Patterns of ethnic enterprise in Miami before and after Mariel. American
Sociological Review, 54(6), 929–949. https://doi.org/10.2307/2095716
Portes, A., & Manning, R. D. (1986).The immigrant enclave: Theory
and empirical examples. In J. Nagel & S. Olzak (Eds.), Competitive
ethnic relations (pp. 47–68). Academic Press.
Prashad, A., Cameron, B., McConnell, M., Rambaran, M., &
Grierson, L. (2017). An examination of Eyal & Hurst’s (2008)
framework for promoting retention in resource-poor settings through
locally-relevant training: A case study for the University of Guyana
Surgical Training Program. Canadian Medical Education Journal, 8(2),
e25–e36. https://doi.org/10.36834/cmej.36849
Premdas, R. R. (1995). Ethnic conflict and development: The case of
Guyana. United Nations Research Institute for Social Development.
R Core Team. (2025). R: A language and environment for statistical
computing (Version 4.5.2). R Foundation for Statistical Computing. https://www.R-project.org/
Rogers, A. (1992). The new immigration and urban ethnicity in the
United States. In M. Cross (Ed.), Ethnic minorities and industrial
change in Europe and North America (pp. 226–249). Cambridge University
Press.
Roopnarine, L. (2003). Indo-Guyanese migration: From Caribbean
homelands to developed nations. In H. P. McD. Beckles & V. A.
Shepherd (Eds.), Caribbean freedom: Society and economy from
emancipation to the present (pp. 583–591). Ian Randle Publishers.
Roopnarine, L. (2007). Indo-Caribbean migration: From periphery to
core. In L. Roopnarine (Ed.), Contemporary Caribbean cultures and
societies in a global context (pp. 189–211). University of the West
Indies Press.
Roopnarine, L. (2018). The Indian Caribbean: Migration and identity
in the diaspora. University Press of Mississippi.
Rose, E. A. (2002). Dependency and socialism in the modern Caribbean:
Superpower intervention in Guyana, Jamaica, and Grenada, 1970–1985.
Lexington Books.
Ruggles, S., Fitch, C. A., Goeken, R., Grover, J., Meyer, E., Pacas,
J., & Sobek, M. (2024). IPUMS NHGIS: Version 18.0 [Dataset]. IPUMS.
https://doi.org/10.18128/D050.V18.0
Ruggles, S., Flood, S., Goeken, R., Grover, J., Meyer, E., Pacas, J.,
& Sobek, M. (2020).IPUMS USA: Version 10.0 [Data set]. IPUMS. https://doi.org/10.18128/D010.V10.0
Sanders, J., & Nee, V. (1987). On testing the enclave-economy
hypothesis. American Sociological Review, 52(6), 771–773. https://doi.org/10.2307/2095552
Sanjek, R. (Ed.). (1990). Caribbean Asians: Chinese, Indian, and
Japanese experiences in Trinidad and the Dominican Republic.
Asian/American Center, Queens College, CUNY.
Sanjek, R. (1998). The future of us all: Race and neighborhood
politics in New York City. Cornell University Press.
Sassen, S. (1991). The global city: New York, London, Tokyo.
Princeton University Press.
Strachan, A. J. (1980). Government sponsored return migration to
Guyana. Area, 12(2), 165–169. https://www.jstor.org/stable/20001322
Thomas, C. Y. (1984). The rise of the authoritarian state in
peripheral societies. Monthly Review Press.
U.S. Census Bureau. (1982). 1980 Census of Population: Users’ guide,
Part A, Text (PHC80-R1-A). U.S. Government Printing Office.
U.S. Census Bureau. (1983). 1980 Census of Population and Housing:
Public Use Microdata Samples technical documentation. U.S. Department of
Commerce.
U.S. Census Bureau. (1993). 1990 Census of Population and Housing:
Public Use Microdata Sample (PUMS) technical documentation. U.S.
Department of Commerce.
Vertovec, S. (2007). Super-diversity and its implications. Ethnic and
Racial Studies, 30(6), 1024–1054. https://doi.org/10.1080/01419870701599465
Wainwright, L. (2012). The emotions and ethnicity in the
Indo-Caribbean. In M. Svasek (Ed.), Moving subjects, moving objects:
Transnationalism, cultural production and emotions (pp. 205–222).
Berghahn Books.
Waters, M. C. (1999). Black identities: West Indian immigrant dreams
and American realities. Harvard University Press.
West, K. K., & Robinson, J. G. (1999). What do we know about the
undercount of children? (Working Paper No. 39). U.S. Census Bureau.
White, M. J. (1988). The segregation and residential assimilation of
immigrants. American Sociological Review, 53(6), 916–923. https://doi.org/10.2307/2095799
Wickham, H. (2016). ggplot2: Elegant graphics for data analysis (2nd
ed.). Springer-Verlag. https://ggplot2.tidyverse.org
Wickham, H., Averick, M., Bryan, J., Chang, W., McGowan, L. D.,
François, R., Grolemund, G., Hayes, A., Henry, L., Hester, J., Kuhn, M.,
Pedersen, T. L., Miller, E., Bache, S. M., Müller, K., Ooms, J.,
Robinson, D., Seidel, D. P., Spinu, V., & Yutani, H. (2019). Welcome
to the tidyverse. Journal of Open Source Software, 4(43), Article 1686.
https://doi.org/10.21105/joss.01686
Wickham, H., François, R., Henry, L., Müller, K., & Vaughan, D.
(2023). dplyr: A grammar of data manipulation (R package version 1.1.4).
https://cran.r-project.org/package=dplyr
Wilson, K. L., & Portes, A. (1980). Immigrant enclaves: An
analysis of the labor market experiences of Cubans in Miami. American
Journal of Sociology, 86(2), 295–319. https://doi.org/10.1086/227240
Wong, D. W. S. (1997). Spatial dependency of segregation indices.
Canadian Geographer, 41(2), 128–136. https://doi.org/10.1111/j.1541-0064.1997.tb01391.x
Zelinsky, W. (1971). The hypothesis of the mobility transition.
Geographical Review, 61(2), 219–249. https://doi.org/10.2307/213996
Zhou, M. (1992). Chinatown: The socioeconomic potential of an urban
enclave. Temple University Press.
