capitals of the top five countries of origin in each year.23 The expectation is
that countries of destination which are geographically closer to the top five
countries of origin in any particular year will attract relatively more
asylum-seekers.
Third, to test the role played by a country's liberal reputation, I used
overseas development aid (OECD data measured in million $) as a proxy
variable for a country's 'liberalness'. I expect a more liberal country (i.e.
one with relatively high ODA/GDP ration) to attract a relatively high
number of asylum seekers.
Fourth, to test network/historic ties theories, I add (at t-1) the stock of
foreign population from the top five asylum countries (at time t). The
expectation is that countries with an already relatively large stock of
foreign nationals from the main countries from which asylum seekers are
originating in a particular year, will receive a relatively greater number of
applications in relation to their size.
Finally, to analyse the importance and the effectiveness of
asylum/deterrence policy measures I use a deterrence index fluctuating
between 0 (lowest deterrent effect) to 5 (highest deterrent effect). To
calculate the index, I analysed two sets of annual yearbooks, the OECD's
'Trends in International Migration' (SOPEMI) and the US Committee for
Refugees' 'World Refugee Survey' for the years 1984-1999. Each describes
and analyses developments in national asylum policy measures for each
country in the paper's data set.
Five measures in particular stand out that have been widely regarded by
policy makers as having the potential to significantly influence an asylum
23 To do this I used a programme developed by John A. Byers which can be accessed at
http://www.wcrl.ars.usda.gov/cec/moregen.htm.
18