Demographic futurity: How statistical assumption politics shape immigration policy rationales in Germany

Demographic rationalities and the epistemology of the status quo

Processes of demographization involve both knowledge production and power relations, as Michel Foucault’s concept of ‘rationalities’ helps us to understand (Foucault, 1982; Lemke, 2007): knowledge about population or demographic problems is constituted at the same time as specific government strategies are evoked that claim to solve (future) crises interpreted as demographic ones. Of course, knowledge production about the population and the administration of the population by the modern capitalist state is nothing new, as Foucault has analysed with his approach to biopolitics focusing on the formation of European liberal political economy. He showed clearly the paradoxical genealogy of the concept of ‘population’, which was constituted as a natural political fact at the same time as it became a central issue for state intervention (2003: 242, 239f; 2009: 83f). Other scholars have elaborated on this idea, and have emphasized that the government of population(s) was constitutive for the development of the nation form (Anderson, 1991; Balibar, 1992). Moreover, the genealogy of ‘population’ is intrinsically linked to the colonial administration of the subaltern and to the history of racialization of (non- or less-) humans, with internal hierarchization and inbuilt selectivities as constitutive for the concept of ‘population’ (Stoler, 1995; Supik, 2014: Tellmann, 2013; Weheliye, 2014). Despite these continuities there have been, on a less abstract level of analysis, different historical phases and settings with different hegemonic approaches towards ‘population’. I suggest that we can grasp the current phenomenon of ‘demographization’ as a conjuncture in which the calculation and administration of population(s) has also become explicitly addressed by governments within the Global North, via a direct reference to demographic data, while other conjunctures have been shaped by more indirect biopolitical approaches via family, welfare, health or migration policies. ²

Demographic rationalities shape political problems and solutions in a specific way. Harvey’s analysis (2001) of the population-resources epistemology, which he developed with reference to Malthus’ principle of population, ³ is helpful as a way of understanding the specific reductionism on which this epistemology is based. In Harvey’s view, arguing that population is the decisive variable explaining social crises and governing the threat of scarcity neglects social inequality, production and consumption patterns within capitalist social relations and the historicity of culture/nature relations. Therefore, demographic rationalities can be characterized by what I call an ‘epistemology of the status quo’. This epistemology involves, as one constitutive aspect, the problems of essentialism within statistical knowledge production in general (Hummel, 2000; Murphy, 2017; Schultz, 2015; Supik, 2014). Demographic rationalities tend to ‘essentialize’ social relations by ascribing fixed characteristics or properties to specific population groups and by introducing reductionist and reifying forms of analysis (cf. Hacking, 1999). On this basis, the suggested solution of any social problem can be found by regulating the size of a (specific) population affected by this problem. For example the number of old people is addressed rather than caring relations, the number of qualified workers rather than education and qualification policies, or the number of immigrants rather than racism become the starting points for problem formulation and strategies designed to produce solutions. This leads to another important aspect: demographic rationalities establish a somehow opaque relationship between ‘quantitative’ average phenomena in their relation to each other (the population, resources and territory as quantitative units) on the one hand, and a selective ‘qualitative’ categorization of difference on the other. For example current demographic rationalities within Europe combine the problematization of an average birth rate or overall net migration with selective criteria when it comes to political solutions. Then, very specific groups are addressed as present in either insufficient (e.g. middle class children or highly qualified immigrants) or excessive (‘poverty immigrants’ or children of black, indigenous and poor women in the Global South) numbers. Certainly, this racist and class-selective differentiation is central to demographic rationalities and leads back to what Tellmann (2013) has analysed as the ‘internal hierarchization of the notion of population’, which she sees as central for the colonial genealogy of liberal political economy. However, it is also characteristic of current demographic rationalities (and is one of the legacies of neo-Malthusianism) that these discursive switches and breaks between abstract quantity and selective hierarchization remain opaque, and are not addressed or legitimated explicitly (Schultz, 2015).

The spatiality of demographic rationalities is moreover shaped by a strong methodological nationalism constituting national populations as if in a national ‘container’ (Hess, 2011; Wimmer and Glick Schiller, 2003). The nation form is inherent to demographic knowledge production because the nation state produces the data through its statistical offices, census activities and registration procedures and also reifies the national container further by producing data on average national birth rates, mortality rates and age structures. Moreover, the methodological nationalism of demography is linked to the idea that a national population reproduces itself ‘naturally’ via fertility and mortality while migration is only shaped as an additional factor of ‘replacement’ (see Lanzieri, 2013: 2). This institutionalized privilege of ‘natural’ reproduction as the imagined guarantee of a nation’s continuity forms the basis for an inbuilt racism within demographic knowledge production. ⁴ This goes beyond the explicit conservative and right-wing demographic narratives within the Global North favouring pronatalist policies promoting higher birth rates among white middle class ‘authoctonous’ women while refusing immigration policies as a central demographic strategy, and touches on how reproduction and immigration are generally put in relation to each other within demographic methodological nationalism.

Certainly, the national container is not the only spatial unit within processes of demographization.

Within the European Union’s policies, there is a certain intent to regionalize demographic rationalities in the sense of a demographic methodological regionalism, which is designed to constitute ‘the’ European people and common European demographic problems and solutions (Eurostat, 2015; cf. Repo, 2016). However, there are also nationalizing trends within European demographic policies, not only because they depend on the aggregation of national data. European demographic expertise often concentrates on comparative approaches, comparing national demographic dynamics, policies and ‘best practices’ (Interview B and C). ⁵ Moreover, within colonial and postcolonial demographic narratives about the Global South the talk centres on global population trends, generally referring to dangerous surplus populations and the carrying capacity of the planet (cf. Bashford, 2014). The biased switch from national to continental or even planetary spatiality reflects the general racializing hierarchies within global demographic epistemologies. Nevertheless, global population data are also based on aggregated national censuses and population statistics which have been established worldwide by the international population programmes.

Last but not least, demographic rationalities are often based on long-term population projections referring to the future in order to legitimate current policies. As population projections depend on past and present statistical data and on assumptions about current trends, they extrapolate the present into the future and so represent another dimension of an epistemology of the status quo. The further into the future these projections are constructed, the more they amplify these assumptions about the status quo and are seen to entail an urgent political need to either adapt to this demographic future or establish explicit population policies in order to change it. This paradox circularity has been described by Elena Esposito showing that statistical knowledge production produces ‘present futures’ which will never happen but are nevertheless established in order to make decisions today and which are producing ‘future presents’ which we will never be able to know (Esposito, 2014; cf. Baldwin, 2012). This production of specified uncertainties and manageable alternatives is moreover not a neutral statistical effect but embedded in powerful economic narratives of future scarcity and postcolonial spatio-temporal difference. Generally, the threat of future scarcity is a foundational element of the capitalist political economy which thereby always already involves the dimension of the future in a specific way, as both uncertain and controllable (Harvey, 2001; cf. Esposito, 2013 with relation to the financial crisis). Demographic rationalities refer to the threat of scarcity in the future, either in the Global South as a lack of resources linked to a threat of overpopulation, or in the current Eurocentric version as the danger of insufficient human capital for envisioned capitalist growth. As Smith and Vasudevan (2017: 211) have argued in the context of critical geographic research, ‘future imaginaries are central for current biopolitical operations’. The demographic temporality analysed here with regard to Germany therefore cannot be separated from a global spatiality which divides current and future geographies. Smith and Vasudevan (2017: 211) explain, focusing on a ‘spatio-temporality of race’: ‘This epistemological grammar centres futurity in the white subject and disqualifies non-human subjects from full humanity and thus from a forward-oriented agency, or confines these subjects to zones of death and sacrifice in service to white futurity’.

Opening the black box of statistical laboratories by analysing situated procedures, methods and narratives about the future is one approach that helps us to understand how the political project of demographization is constituted (Callon and Latour, 1981: 284). For this purpose I refer to those publications within critical policy analysis which have focused on scientific policy consultancy and more generally science policies (Hermann, 2009; Loeber et al., 2005). Only by analysing the concrete procedures of statistical and demographic knowledge production within the concrete settings of statistical offices and demographic think tanks does it become possible to understand how the spatio-temporal epistemologies of the status quo are (re)produced today.

65 or 80 million inhabitants in 2060? Eurostat’s and Destatis’ assumption politics on net migration

Which specific assumptions do the statistical offices use in their population projections, and what data are produced on the basis of these assumptions? A comparison between the most recent long-term population projections of Eurostat and Destatis shows a huge variety of outcomes. Both statistical offices work with what is known as a ‘deterministic’ approach (Lanzieri, 2017; Pötzsch, 2016; cf. Messerschmidt, 2017). They calculate various scenarios, based on different assumptions about future fertility, mortality and migration, but present only one, in the case of Eurostat, or two, in the case of recent Destatis projections (Destatis speaks of a ‘corridor’) as the main scenario(s). Both institutions usually avoid terminology such as ‘forecast’, ‘prognosis’ or even ‘probability’ when reflecting the status of their projections. However, the way in which the results of these main scenario(s) are presented to and interpreted by the public nevertheless suggests that the main scenario(s) are at least probable. For example Eurostat introduces the Europop2013 projection as a ‘what-if-scenario’ to the public, but then uses explanations based on the main scenario like the following: ‘Eurostat projections indicate that the EU population will grow …’ (Eurostat, 2015: 158). The way in which detailed graphs and figures are shown under the general title ‘who are we’ also suggests that the projections are factually objective knowledge (similarly Destatis, 2015).

In the following, above all, I will refer to four population projections, two by each statistical office: for Eurostat the Europop2013 projection and the most recent updates on Eurostat’s website ⁶ ; for Destatis the 12th Coordinated Population Projection (CPP) from 2009 and the 13th CPP from 2015 (2009, 2015). Beginning with Eurostat: comparing the Europop2013 data for 2060 with the Eurostat, 2017 update, the aggregated data for all 28 EU countries together remained quite stable. Europop2013’s main scenario resulted in a European population increase from 507.2 million Europeans for 2013 to 522.8 million for 2060 (EC, 2015: 2017), and the, 2017 update resulted in an increase from 508.4 in 2015 to 524.6 in 2060 (Eurostat, 2017a). These data indicate that the future of a declining European population, which was very present in the 1990s’ and early 2000s’ policy documents, was not confirmed (cf. EC, 2005: 2). However, in contrast to these stable aggregated data, there was a huge change in Eurostat’s national data between 2013 and 2017. For Germany, Europop2013 projected a reduction of the population from 81.3 million in 2013 to 70.8 million inhabitants in 2060 (EC, 2015: 217), but in the 2017 update the German population remained stable at 81.2 million inhabitants in 2015 and 80.8 million in 2060 (Eurostat, 2017a). In other words, there has been a massive change from the thesis of a ‘shrinking’ to a stable German future population. ⁷ In contrast, the last two official projections by Destatis resulted in a continuation of the thesis of population decline. The 12th projection from 2009 calculated 65 million inhabitants for Germany in one main scenario for 2060 and 70 million in the other, and the 13th projection calculated 67.6 million or 73.1 million inhabitants (Destatis, 2009: 5; 2015: 6). Even in the most recent Destatis update, published after the criticisms of its migration assumptions in 2015, Destatis’ projection continued to support the thesis of a ‘shrinking’ German population, calculating 76.5 million inhabitants for 2060 (Destatis, 2017).

This wide range of different outcomes of these four Eurostat and Destatis projections for 2060, ranging from 65 to 80.8 million inhabitants, demonstrates that long-term projections result in very different outcomes depending on the year they are produced and on different assumption politics. Changing and adapted assumptions about net migration had a very strong impact on these different outcomes. This becomes clear when we examine the net migration assumptions within these four projections. Europop2013’s main scenario projected a net migration for Germany of 228,700 in 2020 and 220,200 people in 2030 (EC, 2015: 216), and increased this assumption for the 2017 update to 327,319 for 2020 and 268,069 for 2030 (Eurostat, 2017b). ⁸ Destatis (2009, 2015), meanwhile, assumed in both projections a stable corridor of a significantly lower net migration of either 100,000 or 200,000 persons per year until 2060. The only variation introduced by Destatis in April 2015, when a steadily increasing and higher than 200,000 net migration had already been registered since 2012 ⁹ and could have been predicted to be much higher by the end of 2015, was the introduction of a short transition period from 2016 to 2021, for which the assumptions in both main scenarios were increased. ¹⁰

How do Eurostat and Destatis produce these differing assumptions on net migration? According to Eurostat official Giampaolo Lanzieri, Eurostat assumptions are developed by splitting them into three different time horizons (Lanzieri, 2017). For short-term projections (2 years) Eurostat uses incoming preliminary official data from the member states; this means that the data are being permanently updated. The medium-term projections (until 2050) are based partly on national migration data going back to the 1960s, which are extrapolated into the future. The third strategy Eurostat uses, progressively for the mid-term period and exclusively for the long-term projection (2050–2080), is especially interesting. Here, Eurostat refers to the assumption of ‘convergence’. Using the average development of recent 20 years as reference, Eurostat assumes that the figures for European countries will move closer to each other in the future. For this purpose Eurostat assimilates the demographic factors of different member states to each other. According to Lanzieri (2017: 4), convergence is an ‘assumption of countries moving together and getting – demographically speaking – closer due to the influence exerted by (converging) socioeconomic drivers’. In other words, a normative political agenda of homogenizing the EU is part of the assumption politics.

The way Destatis produces assumptions on future net migration is in some respects similar and in other respects very different from this Eurostat policy, which explains the lower outcomes. Like Eurostat, Destatis refers to long-term trend extrapolations based on an even longer retrospect going back to 1954. In 2015, Destatis calculated average net migration since 1954 as 186,000 persons per year, thus legitimizing the stable 100,000–200,000 assumption corridor (FSOG, 2016: 6). Moreover, the office incorporates projected long-term demographic change within the main European countries of origin (Eastern and Southern European countries) into its calculations, arguing that the main migrating group, the young population, will decrease in these countries and migration towards Germany will therefore also decline. The office also admits that there is a much younger age structure in other countries of origin such as Syria, Afghanistan, or Iraq; however, this does not change their assumption perspective and is interpreted as negligible. ¹¹ These arguments, legitimizing the assumption of lower net migration rates in the future, are further reinforced by Destatis’ officials when predicting more restrictive migration regimes. Already in early 2015, the Destatis official Bettina Sommer assumed that the peak of an increasing net migration had been reached. ¹² Later in 2016, three leading statisticians from the office explained their assumption politics: ‘Policies tend to restrict migration as soon as particular streams gain a certain momentum’ (FSOG, 2016: 8). Moreover, they argued that ‘current policy discussions in reaction to the recent influx of refugees already point to a similar direction’, meaning restrictive changes in immigration legislation (FSOG, 2016: 6). Similarly, a demographer from the government’s Federal Institute of Population Research explained this assumption politics: ‘To a certain degree, population projections involve also calculating how migration policies will probably be in the future. We could also reflect how much immigration we will have when we open the border. Without an idea how we will steer immigration in the future, model calculations make no sense’ (Interview A). In summary, while Eurostat integrates the political agenda of ‘convergence’ into its migration assumption politics, Destatis refers at least indirectly to the prediction of a restrictive immigration policy when designing its assumptions. What is similar and striking in both assumption politics is that the very long-term view into the future is based on a very long-term look back. A closer look shows that this is a very questionable basis, beyond the general question of what current migration patterns have to do with the totally different social reality of 1950–1960s Europe and the two Germanys. One concrete problem with this long-term retrospect is that both Eurostat (2017a) and Destatis (2016b: 22) base their trend extrapolations for the period until 1990 only on the data of the Federal Republic of Germany, ignoring migration to and from the German Democratic Republic. But above all, even when leaving out undocumented irregular migration, past as well as current national data on migration are ‘volatile’, incomplete, uncertain and diverse. In 2009, Eurostat therefore urged the National Statistical Offices (NSI) to use common criteria for migration statistics, resulting according to Herm and Poulain (2012: 161) for Germany in a change in registered immigration from 721,014 people according to the national definition to 346,216 for the EU-mandated definition. ¹³ Generally, German demographers explain a certain reluctance of their discipline to study migration with a ‘very low quality of data’ (Interview D). Despite these serious problems, both offices seem to need these uncertain long term retrospects as stable bases for governing the uncertainty of the future and for being able to present long-term projections at all. Destatis statisticians, sounding rather helpless, explain: ‘It is argued that phases of high net migration always alternated with phases during which the migration situation was calmer. In all likelihood net migration levels will also be subject to great variation in the future as it was usual in the past. Hence, assumptions on future migration levels always have to be interpreted as long term averages’ (FSOG, 2016: 1; cf. Lanzieri, 2017: 3). However, the time scale selected by the NSI for the retrospect is also an aspect of assumption politics with strong implications for population projections and so for the construction of demographic policies, as alternative calculations on less long-term scales show. ¹⁴

Between inevitable and governable futures: The strategic use of demographic markers within immigration policy rationales

The underlying assumptions of population projections have not usually been an issue for public debate in Germany. However, in 2015 diverse actors questioned the Destatis projections. During the ‘summer of migration’ and the enormous increase of immigration to Germany, resulting in a net migration of 1,139,000 persons registered to the end of 2015, a certain public interest in possible demographic effects of migration developed (Schultz, 2016). In this context, diverse voices publicly questioned the net migration assumptions made by Destatis. A number of articles in the media criticized Destatis’ assumptions as too low, as ‘unworldly numbers’ (Lessenich and Messerschmidt, 2015, cf. Messerschmidt, 2017) or as ‘absurd’ (Gersterkamp, 2015). And some think tanks started to produce and publish projections differing from Destatis’ (Deschermeier, 2016; Maier et al., 2016). One of the reactions to the public doubts was thus a pluralization of demographic knowledge production about the future. In early 2017, even the German government in its ‘balance of demographic policies’ compared Destatis’ projections with the calculations of other think tanks, for example of the Cologne Institute for Economic Research (IW), which assumes higher net migration rates for the future (BMI, 2017).

Destatis officials reacted to these criticisms with a mixture of defensive positions, among them referring to those aspects of earlier projections which had been confirmed (FGOG, 2016; Pötzsch, 2016).

They continued to state that their projections were of ‘high relevance’ or ‘realistic from today’s perspective’ (FSOG, 2016: 8). This indirectly reaffirmed the forecast function of projections, and they also rejected the proposal made by some scholars to move to a probabilistic model of calculating population projections, arguing that this would be too difficult for the public to understand (Pötzsch, 2016: 51). ¹⁵ While all these arguments were already common positions in former years, the senior Destatis official Pötzsch added another, somehow reverse type of reaction to the public criticism in 2016. She emphasized once again that population projections are not ‘visions of the future’ but ‘statistically founded extrapolations of demographic structures’ (2016: 3, 39), and also referred explicitly to the function of population projections as what-if scenarios. Population projections should ‘show what would happen if no “integrative” and “constructive” efforts are undertaken. Public policies are able on this basis to shape actual developments, insofar as the projected trends are not desired politically and socially’ (Pötzsch, 2016: 38). ¹⁶ And further: ‘If foreseeable trends can be attenuated or even stopped because of countersteering, reality will necessarily come to differ from population projections’ (Pötzsch, 2016: 38). Moreover, she referred to recent policy reforms such as the raising of the retirement age or parental leave, which in her view ‘would not have been imaginable without knowledge from population projections’ (Pötzsch, 2016: 38). These arguments from the statisticians show that population projections are involved in a paradoxical circle also in relation to how they frame immigration policies. On the one hand, population projections understood as ‘realistic’ forecasts integrate (generally invisible and depoliticized) assumptions about future (in the case of Germany restrictive) immigration policies as probable facts. On the other hand, when population projections are interpreted as what-if scenarios, they also provide a scientific basis for designing ‘demographized’ immigration policies, a circular relationship.

How are these ambiguities of demographic rationalities, suggesting either an inevitable or a governable demographic future, spelt out concretely within current German immigration policies? Within recent German narratives by the government and consultancy expertise, we can see that different statistical markers are referred to when highlighting one aspect or another, which results in implicit strategies of ordering these ambiguities. In the following, I will show three different elements of the discourse on immigration and demography. Before looking more closely at these three aspects of how the ambivalences are ordered, it needs to be noted that the trope of a ‘shrinking’ Germany has not recently been a central issue within the policy consultancy and governmental discourses on demography and migration. Current expertise highlights above all future age proportions as the criterion for successful demographic policies, maybe because of the more and more obvious uncertainty of the ‘shrinking’ thesis. Nevertheless, within media and policy circles alarmist perspectives referring to Destatis’ long-term projections and to the vision of a disappearing German nation remain relevant, and are currently being revitalized by extreme right-wing parties and movements. ¹⁷

Long-term projections of age proportions function as a stabilizing element within demographic rationalities and are referred to in order to strengthen the aspect of inevitability (cf. Eurostat, 2015). Ageing has thus become a secure basis for projection producers and interpreters in uncertain times. For example Destatis declared in 2016: ‘Currently high immigration cannot reverse population ageing’ (Destatis, 2016a). Similarly, the recent governmental balance of demographic policies explained: ‘Current immigration will not slow down the ageing of the population substantially until 2035’ (BMI, 2017: 8). This report refers to the ‘old age rate’, the percentile relationship between the population of pensionable age and of working age, and emphasizes that the old age rate increases from 34.7% in 2015 to 46.8% in 2035 according to Destatis’ projection and, with only a slight difference to 44.9% in the alternative projection of the Cologne Institute (BMI, 2017: 8). The future ‘old age rate’ has always been one of the central elements within demographic policies legitimizing neoliberal pension reforms or austerity policies. Trade unions and critical welfare studies scholars have questioned the reductionist nature of this demographic argument. This narrative ignores increasing social inequality within generations, and is inadequate as a way of understanding the capitalist dynamics of labour markets, wage development, precarization, unemployment and productivity as complex processes underlying the development of public pension funds; it also ignores the broad range of possibilities of redistributive reforms (Bontrup, 2014; Butterwegge, 2015; Cooper, 2016). Referring to immigration policy debates, racist anti-immigration forces have long referred to this calculated low effect of different future migration rates to long-term age relations within Germany – often in addition to their alarmist rhetoric about the threat of a changing ‘population composition’ or even ‘population exchange’ because of immigration. Generally, the argument is that ‘justifiable’ higher rates of future net migration would not change essentially the process of ageing because immigrants are also getting older (Schultz, 2016; Birg, 2001; cf. UN, 2000). More recently, policy consultancy experts have argued that another reason for this is that immigrant populations ‘adapt “their” reproductive behaviour’ to Germany’s low fertility rate (cf. BAMF, 2011). ¹⁸ These arguments are standard elements of demographic expertise, which always states that immigration can at best slow down but not halt demographic change. For conservatives and right-wing nationalists, these calculations serve as an (additional) argument against liberalizing immigration policies, resulting in the argument that the only adequate demographic solution is a ‘self-sustaining’ birth-rate on the part of the Germans (for example the right-wing party ‘Alternative für Deutschland’, see AfD, 2016: 41).

However, those political forces which focus on the future workforce requirements for the German economy and use population projections as what-if-scenarios refer to a different marker. The marker of the absolute number of people of employable age and mostly medium-term projections become relevant here. This marker appears more dynamic in the sense that it varies more depending on different projection scenarios. In recent years, labour market and employer-associated think tanks and political parties have referred to this marker in particular in order to call for utilitarian migration management policies. Nevertheless, the what-if character of these projections is not always communicated, which causes a lot of confusion. One strategy of this type of projections is to publish hypothetical future projections with no net migration at all, in order to emphasize the urgency of political action against the threat of a lack of working-age population in future. For example in 2011 the government’s employment agency, published a study calculating that in 2025 Germany would have a shortfall of 6 to 7 million people of employable age – assuming hypothetical zero future net migration. This alarmist figure circulated widely in the media without an explanation of the underlying assumption (BfA, 2011). Also, the Social Democratic Party, when advocating a new immigration law in early 2015, referred to these hypothetical model calculations and a correspondingly steep decline of 6.7 million in the future number of people of working age without explaining that they were based on an assumed zero net migration (SPD, 2015). Another more recent strategy of employer-associated think tanks, the opposite of the strategy just described, is to assume a higher future net migration in order to strengthen their political agenda in favour of utilitarian migration management. For example, as already shown, the employer-associated Cologne Institute assumes higher figures than Destatis in order to show that higher net migration rates would have an effect on the future working-age population (Deschermeier, 2016).

Nevertheless, when it comes to concrete policy measures a third, class-selective marker becomes central within the current German ‘demographized’ immigration policy debate – the marker of the future number of highly skilled workers, which serves to introduce the trope of ‘qualified immigration’. Interestingly, the what-if scenario about a future shortage of people of working age is not translated adequately into policy measures: there are no calls within the hegemonic debate to open the border generally for more immigration or, more specifically, for a younger population of migrants, as the former reference marker would suggest. Certainly, during 2015, there were some reflections from politicians, think tanks and the media arguing that the ‘sheer number of immigrants’ happening at that time could be favourable for the German economy (Nahles, 2015). However, all suggestions about concrete policy reforms move from these considerations towards calculations about how to counteract a future lack of highly skilled workers (Mediendienst, 2015; cf. Georgi et al., 2014; Scherschel, 2016). The main issue for current debates about immigration policy reforms, in addition to increasingly restrictive asylum legislation, is the pervasive reference to a class-selective ranking system for non-EU labour immigrants, which focuses above all on the criterion of professional qualifications. ¹⁹ However, this argumentative shift or break from abstract average quantitative to predominately class-selective, qualitative criteria defining the population to be included in the nation is not explicitly legitimated or explained within current demographic narratives on migration, and remains opaque (Schultz, 2016). Thus, the German version of demographic utilitarianism arguing in favour of recruiting a highly skilled migrant labour force does not necessarily conflict with right wing or conservative demographic expertise in favour of closed borders and ‘ethnic homogeneity’, but presents itself as combinable within current German policies (cf. Kasparek et al., 2017). In the end, the typical switch within demographic epistemology from abstract quantitative calculations to class-selective hierarchization serves as a racist immunization against that part of the global ‘population’ which is assumed to be superfluous and dangerous.

Demographic futurity, racism and immigration policies: Concluding remarks and further research questions

Demographic rationalities, the increasing influence of which is addressed with the concept of demographization, stand generally in opposite to political projects which demand global social change and question existing structures of inequality. For demographic rationalities perpetuate the status quo via an epistemology that depends on the idea that the population (or certain population groups) are the political variable while social relations, global capitalist resource appropriation and distribution and national/regional policies of space and territory are presumed as static.

Processes of demographization are also affecting current rationalities of immigration policies in a specific way, as I have shown for the case of Germany. Some dimensions have been identified as especially relevant for ‘demographized’ immigration policies: the methodological nationalism at the core of demography frames immigration generally as no more than an additional, flexible stopgap resource for ‘natural’ national reproduction and for preventing future demographic crises. We can speak here of an inbuilt institutional racism of demographic knowledge production. Moreover, the switch from abstract quantitative to selective qualitative (racist and class-hierarchical) arguments – with all the weight of colonial and eugenic legacies – can be identified within Germany’s immigration policy narratives as another typical core element of demographic rationalities. In this paper, I have focused on demographic futurity and opened the black box of statistical knowledge production in order to approach it.

Eurostat and Destatis’ assumption politics about future net migration form the necessary basis for long-term population projection, and thereby influence demographic futurity in a decisive way. Although the existing data available to nation states about migration are unable to grasp and register the autonomy of migration, and although there is no ‘demographic law’ of migration that would make it possible to predict the future, the statistical offices Eurostat and Destatis are developing more and more sophisticated strategies within their assumption politics in order to sustain those long-term projections. These strategies differ from one another and involve hidden political agendas, for example the idea of European convergence in the case of Eurostat and the prediction of more restrictive immigration regimes in the case of Destatis. Moreover, they involve the epistemology of the status quo by referring to past average net migration as the basis for long-term trend extrapolations.

While media, consultancy bodies and governments generally interpret projections as forecasts, statisticians emphasize the function of what-if scenarios and the possibility of ‘countersteering’ at a time of increasing criticism of their knowledge production. In the case of hegemonic debates on immigration within Germany, I have identified strategies of reordering the paradoxical ambivalence of projections as forecasts and as what-if-scenarios. The narrative of an inevitable long-term ‘ageing’ of the national population seems to be important for those political agendas which cling to the necessity of neoliberal reforms and, with respect to immigration policies, emphasize the demographic ‘inefficiency’ of a more liberal immigration regime. The narrative that refers to the statistically more dynamic, absolute number of people of employable age, in contrast, seems to be important for those political agendas which refer to projections as what-if scenarios in order to increase pressure to secure the national labour force supply, among them utilitarian immigration management projects. However, when it comes to defining the type of immigrant population specifically needed for the sake of the national future, racist and class-selective criteria lead to a policy that concentrates on immigration options only for highly skilled non-EU workers, while at the same time more restrictive asylum and humanitarian policies are implemented. Despite all these uncertainties and inconsistencies, the current process of demographization seems to have become established in multiple ways. The absurdly long-term dimension of population projections itself has not been questioned within the hegemonic debate. And no significant voice has been raised in favour of simply abandoning this highly speculative form of knowledge production.

The analysis presented here opens the black box of an apparently neutral and depoliticized statistical knowledge production in a Foucauldian tradition of questioning the politics of truth and of critical policy research. This analysis makes it possible to go beyond a critical analysis of new trends of population policies and immigration management measures which would leave the underlying policies of knowledge production and of problematization uncontested. The critical analysis of a demographization of German immigration policies presented here, however, needs further analysis and debate. One important element to be investigated more extensively is the hegemonic relation of forces within processes of demographization. The analysis of hegemonic projects (as a heuristic concept) struggling for influence could be a helpful methodology here (Brand, 2013; Buckel et al., 2017: 16f.). This would make it possible to distinguish very different political forces which adhere to and promote demographic rationalities. This relation of forces will decide how the mixture of utilitarian immigration management in order to recruit skilled global labour and nationalist policies of exclusion develops in the future. And it will enable us to understand to what extent and how very different political projects converge within demographic national (or regional) agendas insofar as they all subordinate immigration policies to the (phantasma of) national economic interests. And all reject thereby approaches that frame immigration policies as policies of rights or approaches that accept the non-governability and unpredictability of transnational migration. Moreover, such an analysis of hegemonic relations of forces will also provide insights into the extent to which and how different political agendas of a new German population policy relate immigration policies to a pronatalist family policy agenda.

Another issue for future research is the need to assess the scope of the trend of demographization itself. Certainly demographization is only one aspect of the German immigration debate which was especially strong in the aftermath of the ‘summer of migration’. However, strategies of securitization and criminalization are at least as important and have become more influential in Germany since 2016 (Kasparek et al., 2017). And, furthermore, the relation between a short-term strategy of ‘economizing’ migration and asylum policies has to be distinguished from the more long-term biopolitical project of demographic policies. Last but not least, the national and regional ‘container’ of demographic policies has to be transgressed within the research itself. The current nationalist or regionalist project of German and European demographic policies, which has as its main question the (future) lack of human capital but at the same time is permeated by strategies of racist exclusion, has to be analysed as one dimension of global bio/necropolitics. The analysis of immigration policy rationalities can serve as an important way of connecting the analysis of population policies in the Global North and Global South and as an impulse for further theorizing about the racist and class-selective internal hierarchization that exists within the concept of ‘population’.