W. Webb Sprague

1
City mouse, country mouse? Modeling rural/urban
population flows
W. Webb Sprague Department of Demography, UC
Berkeley
Introduction -- Why the Red/ Blue split in the
USA? There seem to be two very different USAs.
One sees itself as part of a world-wide network
of culturally rich urban places, the other is
trenchantly local. One depends on finance and
manufacturing, the other on natural resources and
tourism. One votes Democrat, the other
Republican. And one often lives (with important
exceptions) in metropolitan counties and one
doesnt. What creates this dynamic?
There has been a rural versus cosmopolitan
cultural tension in the West at least since the
Ancient Greeks with Hesiod. Is the split due to a
set of physical features like agriculture leading
to different world views? To circulating
discourses of authority and patriarchy (Lakoff)?
To simple quality of life attributes? To the
machinations of culturally savvy neoliberal
media? To a fundamental cultural dialectic
reflecting a tension between intimacy and
universality?
Results Exposure to rurality does have an effect
on subsequent rural residential choices, whether
accounting for education or not (Note that that
all the tables have a p 0.00) I
think that the rural 79 to rural 96 association
is mostly independent of the educational
variable, leaving a cultural explanation as the
most reasonable. See below tables. Unfo
rtunately, the clustering did not yield any
results that I could interpret. Average s(i) .63
wth k2, then dropped with greater k. Needs more
work.
Results, continued The four groups -- rural ba,
rural noba, cityba, citynoba -- show distinct
lifecourse patterns when graphed. Each bar
represents the proportion in each phase.

• Study Aim
• To test the anecdotally driven hypothesis that
  there is a population circulation among
  countrified and citified places in the USA.
• This project requires that
• we substantiate a difference in country and city
  lifecourses, and
• we create a life course typology of cosmopolitan
  and non-cosmopolitan USA residents, with a set of
  representative lifecourses to use as input to a
  macro population model.

• Method
• Overview of approach
• Use the NLSY79,
• use education and urban rural residence as the
  variables -- education because exploratory
  analysis revealed large effect on migration.
• calculate contingency tables to support
  difference in rural urban migration between those
  exposed to rurality in 1979 and not,
• convert each respondents data to TraMineR
  sequences,
• calculate optimal matching sequence distances and
  cluster into a few groups,
• finally, use representative medoid sequences
  from each cluster as input to macro calculations.
• Notes
• Dichotomised education into whether respondent
  ever attain 16th year of college or not, based on
  the importance of a BA degree for overall life
  chances.
• Gender and race seem to have less effect on
  residence choices than education and region I
  ignored them because of sample size issues.
• Cut off analysis at 1996, because definition of
  rural and urban changes in 1998.
• Caveats
• Only used individuals with no missing years
  between 1979 and 1996, and well defined urban
  rural measurements for all waves.
• Cant be sure if migration behaviour is
  accurately proxied by urban/ rural residence.

Red blue map httpwww-personal.umich.edu/mejn/el
ection/2008/countymapnonlinr1024.png Non-Metro
counties http//www.ers.usda.gov/Emphases/Rural/G
allery/nonmetrocounties.htm
Conclusion Havent quite figured out the details,
but I have substantiated what might be termed a
rural culture effect in terms of concrete
population flows. Georeferenced data would be
nice -- county rurality codes. Record linking
would be even better. Possibly revisit more
traditional models and path analysis now that I
have an exploratory sense of the data.
Population stocks and flows driving Red versus
Blue? Or perhaps the tension is due partly to
cultural diffusion along stable population flows?
This idea stems from my experience doing
fieldwork in Alexandria Oregon, town that
values its country identity. Here I met many
people who migrate between small towns like
Alexandria throughout Oregon and the US, moving
yet avoiding the cosmopolitan centers over their
life course. If there are such laminar strata
of circulation, these populations may lead to
identity stances, as well as cultural inertia due
to domestic cultural reproduction. Thus, Red
Versus Blue may, to some degree, be driven by
population dynamics. But this idea needs to be
investigated more rigorously.
References Lakoff Elder corn stupid thing Fischer
Hout TraMineR Cluster people
Acknowledgements Prof. Jenna Johnson-Hanks, Dr.
Sheila Martin, UC Berkeley Deans Normative Time
Fellowship, and