15.08.2012 Views

Document file:///D|/Export1/www.netlibrary.com/nlreader/nlreader.dll ...

Document file:///D|/Export1/www.netlibrary.com/nlreader/nlreader.dll ...

Document file:///D|/Export1/www.netlibrary.com/nlreader/nlreader.dll ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>Document</strong><br />

Page 63<br />

course, central places must also serve the markets they provide for themselves and for each other, which<br />

already starts to sound a bit more like what I have already described. But can anything like the regular<br />

spacing of centers imagined by Christaller and Lösch emerge from the cumulative processes of Pred?<br />

The initially surprising answer is yes. I have carried out a number of simulation experiments with a<br />

highly stylized economy in which locations are lined up symmetrically around a circle. For each<br />

simulation I began with a random allocation of manufacturing across locations, then let the economy<br />

evolve. For some parameter values, of course, all manufacturing ends up in a single location. When the<br />

parameters are such that several manufacturing centers typically emerge, however, they are normally<br />

roughly evenly spaced around the circle. That is, this linear economy spontaneously organizes itself into<br />

a pattern of central places with roughly equal-sized market areas.<br />

I have a pretty good idea of why this is true, but haven't got it fully worked out yet. Let's just say that<br />

''successful" locations, those that end up with a lot of manufacturing, tend to cast an "agglomeration<br />

shadow" over nearby locations, but that rival centers can thrive if they are far enough apart; the result is<br />

thus a number of centers at a more or less characteristic distance. Some work in progress suggests to me<br />

that the spacing will be more regular, the smoother the initial distribution of manufacturing, with an<br />

almost perfectly smooth initial distribution producing a perfect central-place pattern in which the<br />

distance between centers is determined by the parameters of the model.<br />

All this is for a one-dimensional economy, but I am, as Michael Milken would say, highly confident<br />

that the same model extended to two dimensions would produce a<br />

<strong>file</strong>:///<strong>D|</strong>/Export2/<strong>www</strong>.<strong>netlibrary</strong>.<strong>com</strong>/<strong>nlreader</strong>/<strong>nlreader</strong>.<strong>dll</strong>@bookid=409&<strong>file</strong>name=page_63.html [4/18/2007 10:30:26 AM]

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!