A historical-theoretical approach

79

are present to varying degrees in particular cases, rather than being simply present or absent’ (Ibid.:

848, emphasis added). It thus implies degreeism – a full obtainment is not necessary. From such an ideal-typical or family resemblance conceptual perspective, it is much less of a problem that, for example, (1) one of the (whatever) identified attributes of the modern state is not found in a particular case or (2) the intensity or quality of this attribute does not live up to the ‘utopian’ ideal-typical standard.

non-80

scientific work, but rather, and merely, that the suggested features of the HTA should be considered both completely justifiable and productive. In the HTA informing/organizing this study, the following five procedural elements apply and are celebrated (at varying degrees).

1. The study’s theoretical analysis is historically informed

This foundational element of a HTA implies at least the following two things. Firstly, and most obviously, it implies both (1) a productive attention to time, development, temporal patterns and context and (2) that relevant historical literature is allowed to play an integrative role in the analysis. In this way, following Streeck’s (2012b: 23, 21, 3) general advocacy, a higher premium is naturally put on a

‘longitudinal’ perspective that looks at major objects of study, such as, for example, capitalism, ‘over time’, as opposed to a more ‘static’ view that tends to construe these entities as practically ‘timeless’.

Thus, although there is obviously nothing intrinsically wrong with, for example, more narrow regression-based scholarship – it may (if relevant, possible, etc.) very well be productively utilized in a selective manner to, for example, corroborate a certain argument – such typically diachronic statistical/variables-based/correlation-oriented procedures are naturally less suitable for retroductively theorizing events and constellations through the historically informed examination of macro-causal mechanisms and processes operating over long time spans.⁶⁷

Secondly, the historical dimension of a HTA does not mean it is historical in any formal or technical sense, but, much more loosely, merely implies that the work represents a historically contextualized form of theorization. Arguably, this more loose or restrictive understanding also finds representation in C. Wright Mills’ (1959) advocacy, outlined in his book The Sociological Imagination,

empiricist implicit kinds), the former represents a specific version of historical-theoretical analysis that has been molded in relation to the purposes of this study.

67 See Danermark et al. (2002: passim, 154, 161-5) for a more general discussion of the congruity between critical realism and ‘quantitative’ or statistical work and the difference between, and complimentary of, so-called

‘extensive’ and ‘intensive’ designs/strategies.

81

of a sociology⁶⁸ that is, as Kumar (2015: 268) labels it, ‘historicized’. In this study, this rather soft characterization of the historical leg of the HTA, as work that is merely ‘historicized’ or (as this study prefers to describe it) historically informed, implies three things.

x Firstly, although at times being too polemical, Mills (1959: 50, 20, 34, 74, 71) correctly criticized and rejected the two rival positions of ‘Grand Theory’ and ‘Abstracted Empiricism’

for being, respectively, either (1) marked by the ‘fetishism of the Concept’, ‘addicts of the high formalism of ‘theory’’ and for operating at ‘high levels of abstraction’ or reversely (2) marked by a specialized and narrowly oriented ‘methodological inhibition’ and a ‘usual thinness of result’, which is ‘stuck (…) in what are essentially epistemological problems of method’.⁶⁹ Clearly, a proper HTA, or at least the version found useful for this study, should place itself somewhere in between these two extremes (see also Mills 1959: 124): more historically/empirically informed and circumscribed than Grand Theory, but much less narrow, fetishized with data/methods and skeptical of social theory than Abstracted Empiricism. More specifically, one can for example say that while the LAF approach perhaps leans slightly too heavily towards Grand Theory – although still very far from Mills’ ideal-type category – more recent approaches like Varieties of Capitalism (VoC) or the well-known welfare state analysis of Esping-Andersen (1990), for all their worth, leans too heavily towards Abstracted Empiricism.

x Secondly, as non-specialists in history, macro-oriented social scientists ‘cannot generally undertake (…) primary research in the archives’ (Kumar 2015: 271). For this reason, social scientists are forced to draw on either ‘what are taken to be the best or most scholarly publications by historians’ (Ibid.: 271, although cf. 274) or the work of other historically-inclined social science scholars providing a productive historical context (who are themselves either interpreting insights/findings from historians or drawing on other historically informed social science work).

68 To be sure, although Mills (1959: 18-9, fn. 2) for specific historical and disciplinary reasons focused on

‘sociology’, it is both implicitly and explicitly clear that his argument pertained to ‘the social studies’ more generally.

69 Bourdieu (1988: 775) arguably launches a similar type of critique when rejecting (also implicitly through his actual work) both ‘theoreticist theory and empiricist methodology’.

82

x Thirdly, and relatedly, this condition is natural and relatively unproblematic given that the aim is not ‘to outdo the historians at their own game’ or to examine history ‘for its own sake’, but to deal with the past in order to ‘understand our own world, in its own time’ (Ibid.: 277): in this sense, as opposed to the formal objective of historical work and the practice of some scholars within historical sociology, ‘the past is mainly important (…) for the light it sheds on the present’ (Ibid.: 277, 276).

2. The study seeks to productively roll out and reinvigorate a flexible but primarily macro-level historical-theoretical analysis

This element, which is intrinsically connected to the historical leg of the proposed HTA, in the first instance implies a research agenda that focuses on long-term structures, mechanisms, processes, tendencies, patterns, developments, configurations, etc., that operates and are deliberately studied at a relatively abstract macro-level. Obviously, this intentional focus on emergent macro-level entities and processes taking place over long time-spans necessarily or logically entails a trade-off in the form of a loss of detail, depth, intensity or, as qualitative scholars sometimes puts it, ‘thickness’ (on the natural trade-off between breadth and depth, see Gerring 2001: 24-25, 2007: 48-50);⁷⁰ reversely, focusing on the micro-level (whether through, e.g., a narrow case study, field work or running regressions), which is also simply another legitimate choice, entails a weakened scope and breadth and that less attention is put on context, broader structures, the so-called ‘longue durée’, etc.⁷¹ Importantly, focusing on either the micro or the macro-level (or, of course, something in between) is simply a question of research strategy – the two strategies serve different purposes, have different qualities and are in principal equally valid – and, as argued, it is both meaningful, justifiable and productive to operate with macro-causal mechanisms.

70 For a particular, and slightly problematic, take on the more general question of trade-offs in social science, see Gerring (2001: 23-29).

71 For a discussion of the distinction between ‘extensive’ and ‘intensive’ procedures, see Danermark et al. (2002:

161-5).

83

That said, productive historical-theoretical approaches operating at the macro-level preferably do this in a flexible manner. Flexibility in this sense implies that the macro-situated analysis either (1) as Mills (1959: 34) prefers, is able to at varying degrees ‘shuttle between levels of abstraction’ throughout the study or (2) shows awareness of the possible implications of shifting between units of analysis or somehow provides some initial indications of the structure/direction of relevant formally unstudied micro-mechanisms (which may be studied in a later separate study). By scrutinizing the relevant historical development in more details through a case study looking at the two specific countries of the UK and Sweden and, at one point, trying to theorize the micro-dynamics of state legitimation (or at the very least preparing the ground for doing this more extensively in a possible later work), this study lands somewhere in between the above two scenarios of a flexible macro-model.

Although everyone would obviously ideally like to engage in a careful ‘continual shuttle’

between the macro- and micro-level (Mills 1959: 126), this is rarely possible and not even automatically productive to do in all studies; ultimately, the choice of level of analysis depends to a very large extent on questions of skills, scholarly habits, time, word count, practicalities or more generally ‘heuristic appraisals’, or importantly, as discussed above, the specific ontological characteristics of the objects being studied (and their conditions of empirical possibility). Specifically, besides the above factors, this study’s choice of focusing on the macro-level depends very much on timing; given the particular diagnosis of this study, a higher relative premium should be put on inclusive long-term macro-situated functionalist work that in all modesty seeks to genuinely and legitimately focus on ‘substantive problems’, that is, ‘the major issues for publics and the key troubles of private individuals in our time’ (Mills 1959: 21), such as was arguably much more frequent in the mainstream academic world before the 1980s.

3. The study seeks to zoom in on the essential dimensions of certain entities and deliberately focuses on dominant long-term tendencies, dispositions, trends and ‘conjunctures’ cutting across different types of variations

Put simply, following Streeck (2012b: 23, 22), this study finds it most productive to examine

‘longitudinal commonalities’ – that is, search for and scrutinize a relevant ‘inherent generic dynamism’

and focus on how ‘common dynamics’ in various ways are causally connected to ‘parallel trajectories’.

84

Obviously, this deliberate attention to ‘commonalities’ implies a simultaneous de-emphasis on ‘cross-sectional variations’, such as those notably found in the VoC literature (Ibid.: 21, 22). This strategy of course recognizes all kinds of variations (e.g., organizational, institutional, geographical, national, regime-based differences). But this study’s substantive analysis, its overall research agenda, the critical realist stratified ontology and mechanism-based focus on tendentiality and disposition, and the logic and expected productiveness of the applied Weberian ideal-typification procedure and partly ‘family resemblance’ (Collier & Mahon 1993) perspective – which, respectively, among other things deliberately zooms in on the ‘utopia’ or ‘governing principle’ of a category (Weber 1949: 91) and accepts (and actually expects) that not all of the relevant dimensions are necessarily empirically activated or fully operative in many specific cases/instances due to ‘countervailing’ factors – clearly predispose it to favor the examination of shared cross-cutting macro-mechanisms.

Although the study is overall structured through a descending level of abstraction⁷² (see below), which means that the analysis generally becomes increasingly empirically/historically detailed, contextualized and concrete as it proceeds – culminating in a case study that zooms in on two particular countries showing maximum variation across the relevant macro-mechanisms and processes – it is undoubtedly generally more interested in examining temporally preceding, structural/contextual

‘remote’ factors than more immediate, variable/changeable, agency-specific ‘proximate’ ones (on this remote/proximate distinction, see Schneider & Wagemann 2006: 759; Young 2012: 669).⁷³ This study, in line with the largely implicit productive practice of the LAF approach, focuses on a specific emergent configuration, namely modern liberal-capitalist democratic statehood. And as part of this, the study concerns itself with a specific cluster of countries, namely modern Western European liberal-capitalist democratic polities; countries which despite many ‘proximate’ differences share a similar set of ‘remote’ conditions and many times display corresponding general tendencies and dispositions. In this study, as shall be seen/argued, the overall ‘remote’ macro-level dynamic shared across Western European liberal-capitalist democratic polities – within which more ‘proximate’ historically contingent variations (e.g., welfarist, party-specific or agentic) are analytically embedded⁷⁴ – is connected to the tension-filled functional relationship between legitimatory and fiscal state-crafting.

72 Compare this to Giovanni Sartori’s ’ladder of abstraction’ (see discussion in Collier & Mahon 1993: 846).

73 Natural scientists generally refer to a distinction between ‘distal’ and ‘proximal’ factors or causes.

74 See also more generally Kaspersen (2008: 30-39).

85

Additionally, on the analytical organization of this study: variously, certain concepts, categories, models, typologies, periodizations, etc., shall be rolled out or developed in an ideal-typical manner, selectively focusing (in the Weberian sense) on what is ‘essential’ about a certain phenomenon. For example, when variously discussing and mapping the development of party politics – or more precisely, the historically varying relationship between statehood and party politics – the study is, despite recognized variation, primarily interested in the ‘essential tendencies’ (Weber 1949: 91) of a party system or a party-type and the overall development and pattern cutting across the different nationally specific party systems.

4. The study seeks to productively transcend stifling disciplinary borders and synthesize insights from different thematic specializations

The HTA of this study is underpinned by a deep-seated preference for a genuinely ‘unified social science’ (Mills 1959: 138). As Mills (Ibid.: 142) succinctly points out: ‘[t]o state and to solve any one of the significant problems of our period requires a selection of materials, conceptions, and methods from more than any one of [the] several disciplines’. However this is realized it should ideally imply that the unhelpful ‘departmentalization of social science’, the ‘lazy safety of specialization’ or ‘the idea of distinct ‘fields’’, etc., to some non-trivial extent is buried or transcended (Ibid.: 140, 21, 141).

Moreover, in contrast to Mills’ (Ibid.: 139-42) seemingly optimistic diagnosis of his times of a slow but gradual overcoming of specialization and departmentalization, a concentration of academic specialization and proliferation of disciplines and fields (and sub-version within these) over the last 30-40 years has arguably taken place. For this reason, there is an even greater need for transcending disciplinary boundaries than before.

Of course, although they are intellectually unsustainable, the ‘oppositions between disciplines’

should nevertheless be taken seriously in a Bourdieusian manner (1988: 778, emphasis removed); while most disciplinary divisions are ultimately ‘absurd’, connected to ‘false quarrels’, disciplines nonetheless have a causally important both objective existence, ‘as academic departments, professional associations (…)’, etc., and subjective existence, ‘as mental categories, principles of vision, and division of the social world’, etc. (Ibid.: 779, 778, 777; cf. Mills 1959: 140). Thus, when approaching – and ultimately transcending – disciplinary boundaries, one must adopt a classical reflexive Bourdieusian stance (see

86

1988: 784, 782), which challenges the ‘social and historical determinants of scientific practices’ by in this specific case simultaneously incorporating both (1) the ‘objective’ intellectual unnecessity of disciplinary divisions and (2) the actual subjective ‘experience’ of disciplinary specialization and the implications of this for the molding of research and research-related practices.

As part of pursuing its research agenda, this study integrates various social science disciplines, accumulating insights from especially political science, sociology, philosophy of (social) science, history and economics (as well as their various sub-fields), and selectively synthesizes different types of work – qualitative and quantitative, social-theoretical and more applied. Moreover, the study seeks to go beyond a too strict ‘division into theoretical denominations’ (Bourdieu 1988: 779, emphasis removed), aiming to relatively agnostically and pragmatically integrate insights from different theoretical perspectives and schools (aiming for, as Bourdieu puts it, ‘cross-fertilization’), while at the same time recognizing the core tenets of certain positions and the fact that some theoretical combinations inherently work better than others. While this usually underrated explicit aim and procedure of productively integrating, synthesizing, surveying, accumulating, mapping and rethinking/reworking various disciplines/fields/positions/techniques/arguments/results can never be fully realized – accumulating different work in an integrative manner is an in principle never ending process, always in need of further development and always infused with certain tensions and inadequacies on the part of the researcher⁷⁵– it should be recognized as a time-consuming and actually creative analytical technique, which may potentially serve as a genuine intellectual contribution.

Importantly, while this ideal of transcending (particularly) disciplinary and field-specific borders might seem daunting, it should be absolutely stressed that ‘[a] social scientist need not ‘master the field’ in order to be familiar enough with its materials and perspectives to use them in clarifying the problems that concern him [sic]’ (Mills 1959: 142).

Despite the strictly speaking intellectual unnecesity of social science disciplines/fields, all specific research obviously leans more heavily towards some of these than others, and most scholars have a specific educational profile, skill-set, interest and specialization that steers them in certain directions and thus inevitably narrows the social-scientific scope. Thus, while the most appropriate or inclusive disciplinary/field-specific designation for a proper HTA (in relation to this study) would simply ideally

75 As Mills (1959: 142) points out, ‘[it] is quite impossible truly to master all of the materials, conceptions, methods of every one of [the] disciplines’.

87

be historically informed social research, this study nevertheless by necessity take departure in a much more circumscribed academic arena. At a substantive level, the study primarily takes departure in the two main interdisciplinary sub-fields of political economy and political sociology.⁷⁶ Additionally, history (including the interdisciplinary sociological sub-field of historical sociology) – or, much more modestly, a ‘historicized’ or historically informed social science – and key insights from philosophy of social science are utilized and critically underpin the overall study. As for specifically the case study, selective insights/findings from the economics-oriented sub-fields/studies of business history, industrial/employment/labour relations, business administration, etc., also inform the analysis. In the end, this study may perhaps best be described as doing (or striving to do) historically informed political economy and political sociology (or, stating the same thing slightly shorter and more formally, historical political economy/sociology).

5. The study seeks to explicitly accept and take seriously certain inherent trade-offs with regards to research choices

This banal but underappreciated final element simply implies that (1) although every scholar would like to do and maximize everything – roll out every method, design, perspective and style and operate at all levels of analysis, etc. – this is of course not possible, and (2) because of this, trade-offs, in the form of conflicting scholarly choices, should be fully recognized and incorporated into assessments of scholarly work (cf. Gerring 2001: 23-27, 234). Because trade-offs imply that deciding on or maximizing the quality of one thing tends to rule out or minimize the quality of another possible thing – as is, for example, seen in the choice between providing historical context and running regressions or between running regressions and doing participatory field work – research (and particularly research design) inevitably, as Gerring (2001: 23) points out, becomes a question of ‘prioritization’.

Although doing field work almost inevitably bolsters the findings of a regression scholar, this is rarely feasible or meaningful – and indeed rarely seen – for reasons that has to do with a compound battery of heterogeneous historically varying factors and typically conflicting criteria and choices,

76 While political economy straddles the two disciplines of political science and economics and is formally a sub-field of the former, political sociology straddles the two disciplines of sociology and political science and is formally a sub-field of the former.

88

which are in no way easily or uniformly decided/agreed upon or capable of being unproblematically assessed in a transcendental or commensurable manner (cf. Gerring 2001: 26-7). Importantly, in contrast to both Gerring’s (2001: 29, emphasis removed, 30) unproductive decision to not take into account various ‘considerations of expediency’ or the (at one point expressed) rather unrealistic assumption of King et al. (1994: 13) ‘that researchers have unlimited time and resources’ – a decision/assumption which is integral to both of their methodological advice – this study’s call for a realistic HTA not only takes seriously a much broader repertoire of often conflicting decision-making factors (for example, as discussed, not merely an ‘epistemic appraisal’ but also a heuristic/pragmatic one) but also considers these inevitably integral to both the study’s logic of prioritization and any meaningful academic ‘appraisal’ of its results. Put simply, it is perfectly meaningful, defensible and productive to both (1) engage in a relatively wide and integrative macro-level examination and (2) not do everything or address/tackle every principally relevant question or issue while doing this.

To briefly sum up: the specific HTA that is utilized in this study to examine governance in the context of the historical dynamics of legitimatory and fiscal state-crafting is not only informed by the preceding philosophical-cum-epistemological considerations but also non-exhaustively constituted by the above addressed five properties.