Welfare Capitalism in Post-Industrial Times: Trilemma or Power Over Rents?

ABSTRACT

The structure of ‘post-industrial’ economies is widely held to be problematic for welfare capitalism, because of inherent limits to productivity growth in services compared to manufacturing. The so-called post-industrial trilemma is suggested to allow only two of relative earnings equality, high levels of employment and fiscal balance, and has resulted in the widespread policy belief that greater earnings inequality and welfare state retrenchment are unavoidable. This article challenges the micro foundations of this understanding, that the production of economic value is technologically determined by the physical properties inherent in goods and services. In contrast, we argue theoretically, and demonstrate empirically, that production, allocation and distribution are contingent processes better conceived in terms of ‘power over rents’ with associated externalities between sectors. Our analysis suggests that the post-industrial trilemma thesis may have unduly distracted research from the potential for redistributive politics to achieve sustainable levels of productivity growth, fiscal balance and higher levels of earnings equality.

KEYWORDS:

• Welfare state
• work
• wages
• production markets
• labour markets
• socio-economics

Acknowledgements

We would also like to thank Sami Bensassi, Mark Blyth, Fiona Carmichael, Paul Edwards, Bob Hancké, Jason Heyes, Liza Jabbour, Alison Johnston, Mary O’Mahoney and Matthew Watson for helpful comments on earlier versions of this paper.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Paul Lewis is a Senior Lecturer in Political Economy at the University of Birmingham. His primary research interest is how economic product is generated and distributed and the influence of national institutional systems upon this. His recent research has focused upon the role of productivity in determining wages and he is an advisor to the United Nations Industrial Development Organisation on this topic. He has published in journals such as The Cambridge Journal of Economics, The Journal of Social Policy, Economic and Industrial Democracy and The Industrial Relations Journal.

Fei Peng is a Professor of Applied Economics at the department of economics, school of international economics and trade in the Shanghai Lixin University of Accounting and Finance (SLUAF). He gained his PhD at University of Birmingham. His research interests are productivity, wage and labor markets with a focus on Asia Pacific and Europe, and he has published papers in journals such as Regional Studies, Policy and Politics, Economic Modelling, Economics Letters and Journal of Economic Issues.

Magnus Ryner is Professor of International Political Economy at King's College London. His publications include Capitalist Restructuring, Globalisation and the Third Way: Lessons from the Swedish Model (Routledge, 2002), Europe at Bay: In the Shadow of US Hegemony (Lynne Rienner, 2007) and The European Union and Global Capitalism: Origins, Development, Crisis (Palgrave Macmillan, 2017) (the two latter with Alan Cafruny).

Notes

1 Baumol recognises that (ii) is unrealistic and (iv) only partially realistic but suggests that they are included for ease of exposition of the model. In terms of the conclusions of the model, it does not matter whether capital costs are included, it is sufficient to note that the unit costs of the two sectors will diverge due to the labour-saving nature of the progressive sector. Similarly, it does not matter whether money wages rise at exactly the rate of productivity growth in the progressive sector, rather that wages have some correlation with productivity growth. Assumption (iii) Baumol takes to be broadly realistic in this and later work, which is the assumption that Iversen and Wren change.

2 Kaplinsky’s concept of rent is not reducible to the neoclassical idea of a scarce factor return, in excess of the equilibrium marginal product of the factor because of its extra-productivity in particular circumstances (Alchian 1987: 142). Rather, it arises from control of scarce resources, which, following Schumpeter, may be constructed by firms in coordination with their physical and institutional environment, resulting in returns above an average benchmark (Kaplinsky 1998: 10–3; Kaplinsky and Morris 2001: 25–6). In our view, these complex resources are irreducible to a treatment as factors of production.

3 Change in real productivity = ΔQ/ΔL, change in nominal productivity = [ΔP. ΔQ]/ΔL, where ΔQ is change in the quantity of output produced, ΔL is change in the quantity of labour input used in production and ΔP is change in the value-added-price index of the product, calculated as per unit sales price minus the price of intermediary products used in production.

4 It is possible that identifying such a relationship could still be consistent with neoclassical theory, but this would require that changes in average industry productivity would have to be perfectly correlated with changes in the average level of human capital within an industry. This is not a prediction of neoclassical economics and there is no reason why such a relationship should occur as high productivity workers should be rewarded irrespective of industry worked within (Hildreth and Oswald 1997: 330).

5 We focus upon sectoral level analysis because in the literature this is where the technological propensity for productivity increases has been understood to operate. It is also the level of analysis where comparative productivity data has been compiled. We are aware of recent empirical work in economics, Barth et al. (2016) and Song et al. (2016), which has highlighted the role of establishment and firm level wage dispersion in influencing growing individual earnings inequality. However, these studies do not rule out the role of industry in influencing lower level differentials, nor do they rule out the potential role of rent-sharing between firms and workers, which our framework suggests. By having access to real productivity and value-added pricing data we are able to examine the relative significance of each of these variables upon average compensation at a sectoral level. (We would like to thank an anonymous referee for drawing our attention to this literature.)

6 Although conceptually distinct, independently measuring changes in volume and price raises serious challenges. Real productivity indicates the output volume or value added produced by given inputs at constant prices (the number of widgets produced per person hour). In tightly specified industries, productivity (about 15% of volume series in national accounts) may be measured directly in units of the product produced (Lequiller and Blades 2006: 47). However, the majority of volume series are derived indirectly by deflating value data by price indices, because it is easier to calculate price indices than measure volume changes directly. Especially when multiple different outputs need to be combined at higher levels of industry aggregation, the derived volume levels or changes in volume have to be weighted by their prices fixed at some point in time, or an average of prices over time. This has led critics to challenge whether national accountants can independently distinguish volume changes. Nitzan and Bichler make the point that measured changes in volume depend upon the base year of prices chosen, and these prices, also acting as weights when combining heterogeneous products and services, have to be assumed to be an equilibrium reflection of consumer utility derived from buying the products (Nitzan and Bichler 2009: 125–33). In recent years, national accountants have tried to minimise the problem of outdated prices distorting volume estimations by holding prices fixed for a maximum of one year, calculating volume changes for each year separately, and then chain-linking these changes together. This reduces more glaring distortions, but it does not deal with the fundamental critique: The price structure of the past is central to measuring a volume level or change today and acts as a series of weights for summation. If prices incorporate relative producer power, then volume changes reflect not only physical improvements in output per factor input but also the distribution of pricing power at selected points in time. Hence, volume changes based on the most recent structure of prices are the closest that we can get to an empirical separation of price and volume. The alternative is to conclude that all that may be measured is changes in value. It means that whilst measures of volume and price change are distinct, the measure of volume change necessarily incorporates the relative prices of inputs and outputs in the previous period and therefore cannot be understood as truly independent.

7 (Helwege 1992: 77–80) found that there is a significant overlap between occupations and well-defined industries, therefore pure occupational effects, independent of industry, do not have a clear meaning and should be small.

8 Examples of positive externalities from the public sector include user-producer networks between the national health service and pharmaceutical companies in Sweden (Lundvall 1992), or childcare provision that the Third Way also recognises as enhancing competitive advantage by reconciling employment and the development of human capital with child rearing and the production of a highly skilled future workforce (Giullari and Lewis 2005).

9 Real estate is a major component of cost in the delivery of many consumer services and hence lower rental costs would also act to drive down producer prices.

10 The importance of these two mechanisms was stressed by a former Chief Negotiator of the Swedish Confederation of Trade Unions and State Mediator in an interview on September 27, 1993.

11 Most recent models use the weaker concept of ‘generalised balanced growth’, which requires only that the real interest rate is constant. However, this is still consistent with the Kaldor stylised facts of constant output and capital/output ratio growth, simultaneous with structural transformation (Herrendorf et al. 2014: 878–9).

Additional information

Funding

We would like to thank the ESRC seminar series award ES/M002209/1, ‘Understanding the post-crisis landscape: assessing change in economic management, welfare, work and democracy’ and the LSE hosted workshop, ‘Wages in the EMU’ for supporting the discussion of ideas which contributed to this paper.