62 Chasing the Wind: Assessing Philippine Democracy, Second Edition policies that may otherwise be parcellized in opportunistic ways by various local ofcials. Such an advantageous network for the incumbent governor may also be seen as a practical response to the weakness or absence of a political party that normally assumes the orchestrator role. Moreover, the relative autonomy enjoyed by local ofcials not only vis-à-vis central ofcials but also among themselves in light of the Local Government Code is another problem that can be addressed by this network of incumbent provincial ofcials linked through familial bonds. However, this nding seems to go against the conventional wisdom that family networks that dominate elective positions in their localities are more likely to privatize resources with impunity and disregard services and decisions focused on public goods such as improving their HDI and lessening poverty incidence. Perhaps in this kind of network, political families nd it more cost effective to win elections by focusing on services that improve socioeconomic conditions. The more conventional responses stress the usual privatization of government resources through corruption, providing highly selective private goods for their loyal supporters, or engaging in the outright but expensive and unreliable buying of votes. This network of ofcial provincial incumbents makes possible the calculated use of public resources such that some are deliberately allocated for social services to meet minimum expectations while ensuring that enough is pocketed for the private coffers of the incumbent family ofcials. A proxy variable for gauging the impact of electoral competition on HDI and poverty incidence is tested by the effective number of candidates running for governor. The effective number of candidates refers not only to the actual number of candidates running but to the actual votes received by each, disregarding in the process candidates who receive a negligible proportion of the total votes cast. The study shows that a bigger number of “effective candidates” for governorship has a statistically signicant negative impact on both the HDI and the poverty incidence, resulting in the worsening of both measures. Once again, this result appears counterintuitive to the normal expectation that more competitive elections, that is more effective number of candidates would result in choosing better candidates and better prospects for improving socio-economic conditions. One traditional explanation argues that a more competitive election translates to a more expensive process as both incumbents and challengers are compelled to maximize the use of their resources to ensure victory in light of the great uncertainty of the election outcome. Thus whoever wins in this costly exercise is pressured to recoup one's expenses, looting the public coffer in the process and leaving few resources left for public services. Another explanation draws from “selectorate theory” (de Mesquita et al., 2003) which argues that the smaller the winning coalition of the winning candidate, the greater is the likelihood that the incumbent will use private goods to maintain the coalition.