As James Scott writes, to be able to govern, administrative bodies need to make objects of government legible. Yet migrant persons do not fall neatly into the categories of administrative agencies. This categorical ambiguity is illustrated in the tendency to exclude asylum seekers from various population registers and to not provide them with ID numbers, which constitute the backbone of many welfare states in Europe. Based on ethnographic fieldwork in Norway and Finland, and in Eurostat and UNECE, we study how practices of population registration and statistics compilation on foreign-born persons can be beset by differential and at times contradictory outlooks. We show that these outlooks are often presented in the form of seemingly apolitical software infrastructures or decisions made in response to software with limited, if any, discretion available to bureaucrats, statisticians, and policymakers. Our two cases, Norway and Finland, are considered social-democratic regimes within Esping-Andersen’s famous global social policy typology. Using science and technology studies and specifically “double social life of methods,” we seek to trace how software emerges as both a device for administrative bookkeeping and also for enacting the “migrant” categories with particular implications for how the welfare state comes to be established and how welfare policies come to be implemented. We note that even if all statistical production necessarily involves inclusions and exclusions, how the “boundaries” are set for whom to include and exclude directly affects the lives of those implicated by these decisions, and as such, they are onto-political. This means that welfare policies get made at the point of sorting, categorizing, and ordering of data, even before it is fed into software and other administrative devices of government. In view of this, we show that methods enact their subjects—we detail how the methods set to identify and measure refugee statistics in Europe end up enacting the welfare services they have access to. We argue that with increasing automation and datafication, the scope of welfare systems is being curtailed under the label of efficiency, and individual contexts are ignored.

You do not currently have access to this content.