19
Where do mediators go?
Before we can put any trust in the results discussed above, we first have to address the question
of whether these results are driven by the fact that the mediators select themselves to different
conflict contexts. Do mediators separate themselves according to the contexts of conflicts, in
which they choose to mediate? Do different types of mediators get accepted in different
situations? In Model 1, 2 and 3 in Table 4, we can see the effect of a set of explanatory,
contextual variables on the occurrence of mediation by democracies, non-democracies and
organisations. There are two indications that a selection effect is driving the results. Organisations
tend to mediate in conflicts with shorter duration (DURATIONDYAD). While duration of the
conflict dyad has a negative effect on the occurrence of democratic mediation, and a positive
effect on the non-democratic mediation, these effects are not statistically significant. Likewise,
non-democracies – not surprisingly – mediate in less, rather than more, democratic countries.
While level of democracy is positively associated with the probability of democratic mediation,
and negatively related to the occurrence of organisational mediation, the effects are not
significant. Hence, even though there are some indications of a selection effect, these indications
are so weak that we cannot be reasonably confident that there is a systematic difference between
the situations where different types of mediators mediate.
D
ISCUSSION
This study has shown that non-democratic mediators have a positive effect on the probability of
negotiated settlement, while democracies and organisations do not. Mediation by different types
of mediators (MIXMED) also has a positive effect on the likelihood of negotiated settlement. We
can conclude that non-democracies outperform democracies and organisations when it comes to
bringing warring parties in internal conflict to negotiated settlements. While democracies may be
better equipped than other types of countries to manage their own conflicts (Hegre 2001) or
Convention