In present-day discussions on evidence-based policy making, evidence is often (mis-) understood as objective, quantified data. Development research is oriented to influence policymaking process and to change policy, while policy makers are encouraged to formulate policies based on ‘valid and objective' evidence. Quality of evidence sometimes is put as an end, as if in a market of ideas: best ideas will be bought by policy makers. The idea to make evidence ‘as objective as possible' might undermine the role of politics in policy making. Similarly, poverty data and research may also be judged from its level of objectivity. For instance, the ‘objectivity' in deciding certain criteria for a person or household to be categorised as ‘poor', and at which level of ‘poor' they are at, often become the subject of poverty debates. In addition, using those certain criteria may also indicate whether a government have taken effective measures to reduce poverty. Actors, therefore, cannot always be totally objective or neutral because they have their own interests that might affect their interpretation of data. In addition, actors' varying capacities and resources meaning the arena of contestation of ideas and interpretation is not necessarily a fair one. This paper argues that development actors may uphold their poverty data interpretation because they deem it ‘politically correct' and necessary to reduce the political risks incurred to their organisation by the opposite interpretation. This is exemplified through a case study of a contemporary poverty debate in Indonesia that was occurred after an independent local research and advocacy adjusted poverty data and brought different interpretation from the published official data. The case will be analysed to show how poverty alleviation data are political and that it needs to be challenged by a more diverse development actors.