This is a guest post by David Marsh (Institute of Governance and Policy Analysis, University of Canberra, Australia), Sadiya Akram (Queen Mary College, University of London, UK) and Holly Birkett (Birmingham Business School, UK), as part of the Duck of Minerva’s Symposium on Structural Power and the Study of Business. This post draws on ideas developed at greater length in their article found here. Links to other posts in the symposium can be found here.

There has been a revived interest in the last few years in the power of business. This is hardly surprising given the way in which Governments made significant concessions to the banks in the context of the Global Financial Crisis (GFC). Indeed, most, but not all, empirical studies of the power of business have concentrated on the relationships between Governments and the financial sector particularly in the UK and the US. Is it true, as some have claimed that the power of business has increased substantially, thus undermining the operation of contemporary democracy? Of course, this is, in large part, an empirical problem. However, assessing the power of any group within society is not easy and we need a more sophisticated conceptual framework to address this empirical problem.

Much of the recent literature starts by engaging with Charles Lindblom’s Politics and Markets published in 1977. Lindlom saw the power of business as resting on two pillars: the structural position of business in the economy, which meant that decisions taken by business were crucial for the re-election chances of business; and the capacity of business to manipulate the views of citizens. Nevertheless, he thought that the system remained to a large extent pluralist (at one point he called himself a .4 pluralist), because governments also had significant resources that business needed; in particular it controlled policy decisions and had a legitimacy business lacked. As such, for Lindblom, the role of government in a market-driven democracy was to persuade/encourage business to give up some of its veto power in return for policy gains, infrastructural investment and legitimacy. So, at the core of democracy in market systems were negotiations between business and government.

Lindblom’s work attracted significant criticism, particularly from pluralists, such as David Vogel, who were particularly critical of his structuralism. This critique has been taken up and developed more recently by authors like Culpepper and Reinke, Bell and Hindmoor and Johal, Moran and Williams, who, to varying degrees and in varying ways, contend that in order to assess the role of business we need to conceptualise the relationship between its structural position, the agents involved in the interaction between business and government and the major ideas about how and economy, and indeed a polity, should operate.

However, in our view, too little attention has been paid to how these concepts are defined and conceptualised. We need to specify the structures (economic/political/social) which we see as affecting the role of business. We also need to identify which agents, collective and individual, are involved and how they interact. Finally, we should specify which ideas are playing a role, at what level of generality and how these different ideas at different levels of generality interact. So, are we focusing on broad narratives/sets of ideas like neo-liberalism or, more narrowly, on the ideas of bankers, and what are we claiming about the interaction between the two.

Consequently, the literature on the power of business needs to draw upon the more sophisticated literature which deals with the relationships between structure, agents and ideas. For us, it is crucial to explore a dialectical approach, to three relationships, between: structure and agency; ideas and agency; and structure and ideas. If we adopt a dialectical relationship, we are arguing that agents have interests which reflect their structural position, but they interpret those interests in a way which is mediated by the ideas which they hold/ to which they are exposed. When they act they can, but don’t always, change their structural position, which then becomes the context in which the agents act. So, the process is interactive and iterative. This means that to study this interaction we need a temporal dimension.

It is also crucial for researchers to establish where they stand on the issue of ontological separation between each of the three pairs. So, to take one example, are structures ontologically separate from agents, or merely analytically separate? This may seem an arcane point, but it is crucial, because, for most analysts, if there is no ontological separation, so they are co-constituted, then structure, and indeed agents can’t have independent causal power. Interestingly, and rather perversely, Bell and Hindmoor acknowledge ontological separation, but deny independent causal power, thus, in our view, privileging agency, as a result of their focus on bounded-rationality, and ideas, as a result of their constructivism. In contrast, Culpepper and Reinke emphasise the importance of structure, but pay little attention to agency or ideas, although Culpepper focuses on agency elsewhere. Finally, Johal, Moran and Williams, in examining the relationship between the Government and the banks in the GFC in the UK, drawing on Foucault, see capillary power as a way in agents were unconsciously disciplined in ways which benefited the financial sector, even if that project was ultimately unsuccessful. As such, they recognise that ideas can have an influence on agents of which they are unaware. They however give a limited role to agency, given these ideas constraint agents in a major way. What we need is an approach which empirically examines each of the three relationships, over time, but acknowledges that without agents recognising and articulating it, their actions may be shaped by structures and ideas.

We need to raise one other important issue here. If we posit capillary power, or indeed a crucial role for habitus, it immediately raises methodological questions. How do we establish that such power is in operation, given that, by definition, it cannot be directly observed? This is an important issue, but it is the subject of another paper.