Positive Feedback in the Political (Pierson’s Path Dependence)


(From the Open House Project blog.)

I’m reading Politics in Time by Paul Pierson (link), and am struck by how little academic political science seems to affect government policy and political discussion. I find political and social analysis incredibly stimulating, especially given how tiresome I find the current presidential punditizing.

I’m particularly interested in Pierson’s purportedly novel conception of how political institutions develop over time, apparently filling the gaps that other models fail to address. (He sets his conceptions against "historical institutionalism" and "rational choice theory".) His analysis is abstract enough to be rigorous and challenging at first, but takes a broad enough view that he can abstract common elements out of disparate systems in a useful, applicable manner. He seeks to "explicate different ways in which things happen over time in social life, drawing attention to processes that are unlikely to be visible without specifically addressing questions of temporality" (p. 10). (more)

In reading just the first few chapters, I’m surprised at how well the concept of "path dependence" maps onto congressional reform efforts. A concept I would probably have referred to as a "positive feedback loop", path dependence is self reinforcing behavior — development whose onset disproportionately influences further development. Basic examples that come to mind are both debt and wealth, which tend to feed off of themselves.

Pierson applies path dependence to economic theory, illuminating situations that may be well explained by examining self reinforcing structures. For example, various national economies develop divergently, and, rather than taking advantage of each other’s successful strategies, produce very different situations. "Once in place, institutions are hard to change, and they have a tremendous effect on the possibilities for generating sustained economic growth. Individuals and organizations adapt to existing institutions. If the institutional matrix creates incentives for piracy, North observes, then people will invest in becoming good pirates. When institutions fail to provide incentives to be economically productive, there is unlikely to be much economic growth." (p. 27)

Pierson argues that the political sphere is particularly subject to self-reinforcing behavior (aka positive feedback, path dependence, or increasing returns). He outlines four mechanisms that render the political particularly influenced by whatever the current state of affairs is. They are: "(1) the central role of collective action; (2) the high density of institutions; (3) the possibilities for using political authority to enhance asymmetries of power; and (4) its intrinsic complexity and opacity… Each of these features makes positive feedback processes prevalent in politics". (p. 30)

Each of these mechanisms seems to easily map onto Congress in a useful way.

Collective action problems make feedback loops because both politicians and constituents (or any political actor) are largely unable to act alone, and must constantly assess the winning strategy, and what is perceived as the winning strategy. This privileges existing organizations, giving established parties, coalitions, and institutions the distinct advantage of clearing the first hurdle of viability. When effectiveness can be found in groups, and groups are hard to form (and political organizing is insufficiently agile), then those groups’ existence will tend to exhibit self reinforcement. The Internet, and blogging, however, are a productively destabilizing force, giving ad-hoc coalitions and unproved institutions an equal voice, where reputations matter less than well sourced convincing arguments. The Internet also reduces the amount to which political activism involves collective action problems: there is a rather low barrier to participation (digital divide notwithstanding). Broader participation and competition means greater alternatives, leading to more agility and easier transitions, meaning we’re less likely to stay stuck on some self-reinforcing pathway.

A dense realm of institutions similarly exists around Congress and the federal government; they sort of approach being the essence of the institution, the defining meta-institution, comprised of departments about departments, creating the conditions for all other institutions. With such far-reaching work, this complex of institutions will be justifiably risk-averse, weighted down by the seriousness of their task, and the high price of failure. The sheer mass of institutions at play gives reform a much higher cost (and renders them path-dependent.)

Third is the "possibilities for using political authority to enhance asymmetries of power." Congress is full of power begetting itself, as is government generally. Societal expectations and checks and balances are supposed to help define the terms of the equalibria controlling this power. The legislative and executive periodically switch in dominance, as do the parties. The incentives created by elections are enhanced by an information-empowered electorate, helping to reign in self-reinforcing political power structures.

The last political mechanism of path-dependence is "its intrinsic complexity and opacity." Complexity and opacity make political institutions and agents less susceptible to any societal pressure, which is more likely to be mediated through sympathetic agents (the media, lobbyists). While complexity is often necessary, and has a high cost of shedding (see #2, institutional path-dependence), it can be countered by information availability. In other words, while Senate procedures provide an effective shield against criticism for questionable votes only as long as those procedures are hard to explain. Insofar as Congress is inscrutable, it’s less likely to feel real pressure, and more likely to reinforce itself. Insofar as the Internet helps make Congress scrutable and transparent, pressure becomes productive.

I’m looking forward to going through more of Pierson’s research, finding it similarly helpful to Harvard’s Transparency Policy Project or Congress’s own best attempts to survey itself.