Geography Reference
In-Depth Information
Box 31.2 continued
Wallace defined two indices to characterise that
dislocation in a given Health Area. These indices are
formally defined elsewhere
(ibid)
but, in brief, the poverty
index (denoted x
1
) is a community-wide measure of
deprivation, while the community disintegration index (x
2
)
is a measure of existing social support networks. These
indices served as the independent
(x)
variables in a
series of least squares multiple regression models in
which cirrhosis deaths (y
1
) and intentional violent deaths
(y
2
) variously formed the dependent (y) variables.
The regression results, which are summarised in
Table 31.1, yield two critical insights:
1
The process of population displacement served to
strengthen the geographical relationship between
HIV-related pathology and social dislocation.
Models
1 and 2 relate deaths from liver cirrhosis to the two
independent variables for a period immediately prior
to the main phase of population relocation (1970-73,
model 1) and a time period after the main phase of
relocation (1978-82, model 2). Between these two
periods, accounted variability in cirrhosis deaths (as
judged by R
2
) doubled from 28 per cent (1970-73) to
57 per cent (1978-82).
2
The process of population displacement was
associated with an absolute rise in the mean level of
HIV-related pathology.
Models 3 and 4, which relate
intentional violent deaths in 1970 and 1980 to the
independent variables, reveal an approximate
stability in accounted variability. However, inspection
of the Student's
t
statistics associated with the
intercept coefficients indicates that this parameter
made a significant contribution to the 1980 model
but not the 1970 model.
Taken together, these results suggest that the social
sequelae of population displacement in the South Bronx
during the 1970s resulted in both (1) a geographical
focusing and (2) an intensification of two correlates of
substance abuse, the predominant route of HIV
transmission in the borough. It appears more than
coincidental that the same social sequelae should also
be highly predictive of the geographical pattern of the
borough's early AIDS cases (Table 31.1, model 5). An
inference from these findings is that the historical spread
of HIV in the Bronx was geographically focused and
fuelled by the social dislocation engendered by an
iniquitous policy of planned shrinkage.
associations of poorer health and social
disadvantage. Although the case study does
illustrate the difficulties of extrapolating from an
area to an individual, the findings do not, as they
stand, give clear guidance as to the direction of
health and social policies. For example, should
policies be targeted at localities and environments
or individuals and households? The case study in
Box 31.2, set in the Bronx district of New York
City, illustrates the sophistication of spatial and
statistical analysis now used in much health
geography research. The study also illustrates how
a socio-spatial analysis of recent historical data can
give guidance to future public health and urban
redevelopment policies in large urban settings.
Such policies are likely to become ever more
pressing in the mega-cities of the world.
