*The paper discusses the choice of exchange rate anchor policy in Sub-Saharan Africa (SSA). Policy makers have the choice between promoting external competitiveness using a real anchor , lowering the burden of external debt using a nominal anchor or balancing both objectives by using a policy mix of both anchors. We observe that these countries tend to have a nominal or a policy mix anchor. Theoretically, we solve a state space model to explain the determinants and the strategy behind this policy. We find that the choice of policy mix is a two strategy steps. Firstly, authorities would choose the nominal anchor according to the velocity, openness, debt, pass through degrees and exchange band target. Secondly, they would choose the real anchor according to mainly their trade integration with regards to the intervention in the exchange rate market. We estimates our theoretical finding using an MS VAR estimation of these determinants with regards to the basket of implicit peg in SSA country.

**To achieve larger visibility and range of actions Non Profit Organisations (NPO) which fulfil same or similar goals can gather in networks. The network is itself a NPO with a board, a budget and a general assembly, it is equivalent to a NPO of NPO (NPO is understood here as an "association" under the french 1901 law). Most of the time, to achieve a greater visibility the network has to harmonize the practices among its members by imposing norms ("the good practices"). Unfortunately this harmonization leads to numerous conflicts which can end up with split in network (less stability) and/or large leaving among its members (less representativeness). In this paper, taking ground in a previous robust model designed to replicated evolution of NPO networks around conflicts on norms,  we test for the ability of different modalities of communications around norms to create both more stable and representative NPO networks. We test for three levels (who is reached) and  four types (how it impacts agents) of communication on norms. We show that the impact of the communication on norms is the most efficient for NPO networks engaged in complex norms negotiation (with lot of dimensions to norms). Secondly we show that communication should be broad and non aggressive to be able to improve both stability and representativeness. Restricted to the board the communication is only able to improve stability but at the price of less representativeness.