There has been an ongoing debate among researchers, policy-makers and development partners in low-income countries on whether and to what degree non contributory social transfers should be targeted to the poor or paid out universally to every citizen or to all citizens in a particular category. This paper critically discusses the assumptions behind the political economy arguments of targeting and tests whether a universal mechanism is bound to politically excel in a low income country context. A number of authors have argued that going universal is a win-win situation, both for the poor, the middle class as well as those who are in power. We would therefore expect broad-based support behind a universal scheme, in particular in countries where poverty is widespread and targeting also proves administratively challenging. On the basis of attitudinal surveys with the urban, rural and student population in Zambia, we actually detect more support for targeting the poor than the political economy models would predict. These findings are corroborated by experimental evidence from rural Zambia. We discuss the assumptions of the political economy models in the light of these findings and contemplate on potentially decisive parameters that the models currently do not incorporate.