REAL
Keystone sector methodology applied to Portugal
In a global view, pointing out the actors that appeared more than once in the
whole network, we have University Student Association, Municipality, Public
Employment Center, Public Health Center, Private Construction Firm, Middle
School, Hospital, Church, Political Party and University as the strategic entities,
which is mostly consistent with our hypothesis. Through eight public institutions and
one private firm in the construction sector, the entire social network is connected.
Attending that this is the pattern registered during the dictatorship, centralization is a
constant across time that could be pointed as a probable explanation for the
development delay. We will turn to this point, in the comparative conclusions. The
third test is for Efficiency, computed on the basis of these two others, and shows a
pattern of increasing efficiency (see Fig. 1).
4.2.3 Peripheral entities or isolates
These should be the entities least likely to be strategic in the network, once they
are the isolated actors. They are the actors with very low values for local and global
centrality (more than 1Std below the means). There are no peripheral entities in the
ORGANIZATION matrix, though we can consider as peripheral the traditional Club of
old textile aristocracy, together with center political parties and one of the Teacher’s
Union in what Money matrix concerns. Information matrix showed a sufficiently dense
network but 6 entities have values below the 1Std+Mean level: four firms from
agriculture, development, textile mechanics and automobile retailer as the above-
mentioned club along with one of the churches.
The branches are consistent with the declining sectors in economic activity and
its social effects. The political parties that are not in the local power have also low
centrality.
4.2.4. Keystone entities
In a preliminary analysis, TT city showed a very interconnected sample of
entities with no clubs or cliques, which doesn’t give any expectancy for a particular set
of actors displaying unique patterns of relations. Nevertheless, closeness and
Betweeness tests revealed important differences among certain entities that required
promoting the ‘keystoness’ test in order to check for significant changes in the network
structure due to the excision of those entities (called the Cut points).
18