The epistemic case against radical interventions
Radical interventions generate large-scale unintended consequences but the expert community relies on flawed methods that misleadingly suggest the opposite.
Why is this post important to read? Is it because we care about democracy and democracy promotion worldwide? Perhaps. After all, democratic state-building fits in well with the democratic principles that have elevated the US to a higher moral standing in comparison to its authoritarian opponents. Still, a historically grounded appraisal of US foreign policy would suggest that the US government has used democracy promotion to justify interventions to drastically alter and control the political systems of nations such as Afghanistan, Iraq, Kosovo, Libya, and Liberia. This supposedly fervent ideological dedication to spreading democracy worldwide is contradicted by America’s historical record of supporting the most ruthless and violent authoritarian regimes in Latin America and Southeast Asia.
Nevertheless, the post you are reading is important because it makes another key point: no matter what the true motives behind democratic state-building have been, the key problem I am stressing is that the radical interventionism (for democratic state-building) is presented as a feasible strategy.
In a recent article of mine, I presented an epistemic critique of this strategy, and through this critique, I had the opportunity to publish a general critique of any form of radical intervention that aims at transforming a human system from the top down (Trantidis 2022).
Democratic nation-building is advertised upon the idea that democratic institutions lead to political stability and economic prosperity. The expectation is that, by establishing these institutions in a foreign environment, these same outcomes will occur in new environments, provided the plan is executed correctly. This expectation,I argue, relies on a dominant epistemological paradigm in social research which has been both powerful and flawed: that there are linear input-output relationships connecting socioeconomic conditions with outcomes and these relationships can be inferred by statistical analysis of past samples and interpreted as probabilistic statements about how other systems - other countries - will behave. This has been the dominant "transition paradigm.”
The origins of this paradigm of can be traced in the so- called modernization theory and the New Institutionalism, whose foundations are rooted in positivism. Positivism can only be convincing, and can only thrive in a social context, under two ontological assumptions: linearity and ergodicity: under linearity, processes of social and institutional change have identifiable input-output relations that can be measured as statistical variables and, under ergodicity, observations from past samples can be used to plan new interventions and predict their effects elsewhere. Human societies are presumed to be mechanical - similar to thermodynamic systems in physics- and therefore some engineering on them is possible.
Based on these positivist premises, modernization theorists have claimed that certain socioeconomic variables are conducive to democratization and development, meaning that countries must invest in developing these variables and will then follow a similar path of progress with the countries currently enjoying these conditions. New institutionalism later emphasized institutional reforms as the way by which political actors would speed up a country's pathway to democracy and prosperity by incentivizing and directing collective behavior towards achieving these favorable conditions. But, in reality, large-scale interventions have often generated serious unintended consequences whose type and scale were not predicted beforehand, as the developments in Iraq and Afghanistan demonstrate.
In short, we tend to see the human environment as a terrain that can be re-organized by planned interventions guided by expert advice. According to this paradigm, we have to reconfigure the relevant qualities and properties of the system as suggested by empirical (mostly statistical) analysis. If the same variables appear in another place, and no previously unaccounted factor comes up whose effect was not present in the sample and could not be calculated at the time of that analysis, the probabilistic projection of input and output is expected to hold. So, if we know what has worked successfully elsewhere, we can then prompt and accelerate a similar process in another system by changing the relevant institutions and behavioral norms as indicated by previous observations. Maybe a few careful adjustments to local conditions are needed, and that’s it.
This paradigm dominates politics, policymaking and the expert community. Previously, Peter Boettke and I identified this ‘epistemic entanglement’ between theory and policy as the source justifying policy attempts for radical and large-scale interventions to guide society towards a particular direction as feasible (Trantidis and Boettke 2022). Several other policy interventions, from macroeconomic policy to education and public healthcare, are based on this positivist epistemology which, in turn, depends on the ontological premises of linearity and ergodicity. Yet, like democratic state-building, planners and policymakers also come across serious and unpredicted side effects. Why?
I contrast the linear transitions paradigm with a complexity framework that understands social change as a non-linear and non-probabilistic process. Insights from the study of complex systems challenge the epistemological foundations of the linear paradigm and better explain the occurrence of serious unintended consequences from planned large-scale interventions. Change in human systems emerges from continuous interactions of social actors that generate novel and discontinuous behavioral patterns, leading to unique and differentiated trajectories across subsystems of interactions. Understanding this behavioral dimension of complexity sheds light on how human systems evolve, often intermittently and discontinuously.
Regarding democratic state-building, for instance, which my article was about, complex interactions take the form of social reactions such as counterinsurgency and networking that unsettle the local system. Rather than giving the system a desired new direction, large-scale and radical interventions, such as military invasion and state-building, will tend to trigger varied and random sequences of behavioral responses, ultimately increasing uncertainty in a system rather than reducing it. The larger the scale of intervention, the greater disruption and uncertainty it will generate.
Although there are cases in which institution-building appears to have succeeded, these “success stories” have multiple case-specific explanations. They will be wrongly taken as a set of transferrable directions of what should be done or what must be avoided in order to attain a similar outcome elsewhere.
While the rationale behind democracy promotion and democratic state-building have been thoroughly debated and problematized in the academic literature, the ontological and epistemological underpinnings of radical interventionism remain dominant and unchallenged.
A great part of my work has been to emphasize that social change is a non-linear and non-probabilistic process, and requires a complex transitions framework to better understand it. This has come at some considerable cost for my career. Unfortunately, social science research is trapped in a paradigm that enables it to claim relevance and impact, by ‘telling us what we should do’. Unless you play this game, your work does not receive any funding, and is often ignored or marginalized as having no impact and little analytical sophistication.
Yet I insist. Because a closer look at societies as complex systems illustrates why the dominant epistemic paradigm is misleading and dangerous. In human systems, responses to exogenous interventions involve sequences of multiple and intractable behavioral reactions – almost in the form of sequential reverberations - that develop through continuous feedback processes and generate novel and diverse behavioral developments. The scale and nature of these reactions cannot be predicted beforehand. Complexity suggests that, even if the formal institutional order is modified, the behavior of the entire system will most likely not be reconfigured as planned, especially if the exogenous intervention is radical and abrupt. For example, behavioral adjustments may manifest themselves as cascades of behavioral adaptations, such as herd-like reactions, that will completely derail the original plan, leading to very adverse directions, and causing system-level unintended consequences and unforeseen turmoil.
References
2022 – Fallacies of democratic state-building. International Studies Review, 24 (4), first published online, October 2022
2022 – Macroeconomic policy as an epistemic problem (with Peter J. Boettke). Journal of Public Finance and Public Choice, 37 (2): 211-231