154 Chasing the Wind: Assessing Philippine Democracy current flows along a political-electoral reform path—seeking to promote system change by seizing power from a corrupt traditional political elite through political-electoral means. The other flows along a social reform path—aiming to promote change by exercising citizenship power in state policymaking and implementation. The fault line that Franco observes could be traced to an even earlier period when divisions, chiefly ideological, existed among proponents of community development from the 1950s. (Cariño 2001) An NGO leader offers another perspective as to how CSOs are divided into four different groups across two axes—between people empowerment and leadership change, and between engaging the state or steering clear of it: One axis is really about working with communities, organizing and mobilizing them (people empowerment) and of focusing on leadership change through elections, formation, training (leadership change)…a second axis is about “engaging the state” and “not working with any of the state agencies”. (Macasaet 2011) In explaining this division, Macasaet categorizes CSOs into four clusters: 1. 2. 3. 4. People empowerment—state engagement People empowerment—non-engagement with the state Leadership change—state engagement Leadership change—non-engagement with the state These two characterizations of the fault lines or divisions within civil society mirror the common dilemmas among CSOs that should be resolved by reflecting on the wealth of experience that CSOs have had in the last 25 years. The first dilemma pertains to the decision of whether or not to engage the state, or by extension, hold the state, or its leaders, accountable. The second is with regard to the link between leaders of CSOs and the broadest constituency—the mass public. Engaging the state With regard to one dimension of the first dilemma—engaging the state, the divergence of views exists, as Franco stresses, even among those who hold the popular view of democracy (the people empowerment-state engagement part of Macasaet's quadrant). This fault-line, however, seems to have been patched after the split of the national democratic movement in the early 1990s and as some CSOs crossed over to “political society” by organizing party list