The intervention of local elites is often cited as an impediment to policy implementation in many developing countries. In this paper, we present initial results from an original primary household dataset from eight tehsils of rural Punjab, Pakistan. We examine descriptive statistics on patron-client interaction and correlations between household characteristics and that relationship. The study raises some key findings. First, households report connections with a range of officials; they interact most heavily with local officials, but a large number of households also report interacting with their provincial and national politicians. Second, many households report receiving active assistance both from local officials and from provincial and national politicians in accessing certain state services, in particular in applying for national identity cards. Third, households report links with many patrons outside their own biraderi or clan. Fourth, vulnerable households, such as landless and female-headed households, appear less likely to interact with and less likely to receive assistance from patrons, suggesting that patronage activity could increase the inequality of outcomes. Fifth, better-off households appear more likely to assist patrons in a range of areas. Finally, local officials and politicians had tended to recommend candidates in the last election, and rural households were strongly convinced that their vote was not secret from their patrons or officials. This is possibly consistent with patronage-based politics and bloc voting.Keywords: Patron, client, rural, Pakistan.JEL classification: P16, D7.1. IntroductionThe activity of local elites, such as powerful landlords, is frequently cited as a major challenge to social and development interventions in rural areas of many developing countries. Local elites are sometimes said to block efforts at rural development or social assistance because their power is threatened by others helping "their" villagers. Alternatively, they are described as diverting the benefits of social or government programs to those in their families or social networks (elite capture). They may also play a role in wider distribution while reaping the personal or political benefits of acting as a benefactor in securing goods and services from donors or the central government on behalf of their clients. Rural Pakistan is thought to be a prime example of such relationships: indeed, it is even frequently described as a "feudal" society. Analytical work in political economy substantiates these concerns: Cheema, Mohmand, and Patnam (2009) have demonstrated the importance of local elites in this context, and the persistence of their power to resolve disputes and mediate between citizens and the state from the colonial period to the present.However, donor and government programs often fail to take this apparently very significant feature of rural societies into account when designing interventions. For example, Mansuri and Rao (2004), in their critical review of community-driven development initiatives, point out that the question of how the "community" is represented in decision making is critical and underscrutinized. In general, the decisions taken are highly subject to local power dynamics that might completely escape the funder's attention. The relationships between local elites and other community members may, however, play a major role not only in these types of decentralized programs but also in the de facto distribution of more centralized state programs through patronage networks of interaction between political actors and local elites.Conversely, and potentially harder to observe, patrons may play a role in wider distribution while reaping the personal or political benefits of presenting themselves as benefactors providing goods and services. Thus, interventions can affect local power dynamics in ways that donors and central governments might not be sufficiently aware. …