Rethinking Democratization in the Philippines: Elections, Political Families, and Parties 63 In contrast, the larger the winning coalition in support of the winning candidate, the more likely that the leader will use public goods to satisfy the coalition. In short, if the winning margin/coalition is small such as in highly competitive elections, there is greater pressure for the leader to use private goods to maintain the loyalty of the supporters. In contrast, wider winning margins and bigger coalitions in support of the leader which is more likely in non-competitive elections might lead to the use of more public goods for everybody. This behavior may arise when the leader feels more electorally secure and the provision of private goods for favored electoral constituencies and interest groups becomes less urgent. Two related variables focus on incumbent governors having relatives holding national ofce and situations where these relatives in national ofces have held ofce in the past two terms. A governor with relatives who are also incumbent nationally elected leaders does not impact signicantly on either the provincial HDI or its poverty incidence. On the other hand, a governor with relatives who had held national ofce in the past two terms has a signicant positive impact on the improvement of the HDI but has no impact on poverty incidence. Another related variable has to do with the “persistence factor” or the number of years that the incumbent governor and the family hold power in an unbroken cycle. The test shows that a longer “persistence” cycle has a statistically signicant positive impact on the provincial HDI but with no impact on the poverty incidence. These three variables suggest that national resources received by provincial governors may help in addressing socioeconomic issues if done in a sustained period of time. The political “persistence factor” seems to be a logical and empirical requisite for effectively addressing oftentimes intractable poverty-related problems which typically demand many years of sustained growth and redistributive efforts. The challenge is to determine how long a period of the infusion of an outside national resource is necessary to more effectively address development outcomes. In this regard, the impact of term limits on the persistence factor for both national and local ofcials also needs to be examined more closely. A political variable examining the effect of an incumbent governor running under the president's party also shows no statistically signicant impact on either the HDI or the poverty incidence of the province. The normal expectation in this case is that the incumbent governor will have easier access to presidential resources that may be used potentially to address socio-economic problems. This nding suggests, however, that access to presidential resources and prerogatives by an incumbent governor is determined not so much by formal party afliation with the president but by more particularistic criteria such as kinship ties and history of friendships and informal alliances with the executive. Finally, an incumbent governor running as an independent may nd more difculties addressing poverty issues since this variable has a statistically