Conservatism is a rational response to epistemic uncertainty

post by contrarianbrit · 2022-07-31T18:04:36.714Z · LW · GW · 11 comments

This is a link post for https://thomasprosser.substack.com/p/how-conservatism-became-more-reasonable

Contents

  1.
  2.
  3.
  4.
  5.
None
11 comments

1.

Recurrently, humans respond to change in a conservative manner, established traditions protecting against unsettling developments. To a great extent, the reasonableness of conservative responses hinges upon external conditions, change being more rational in certain ages.

Today, we may be living in an age in which conservatism is more reasonable. This reflects increasing doubt about the ability of politics to achieve change. Following a replication crisis which suggests that many academic results are spurious or overestimated, there are fewer reasons to trust policy interventions. 

Relatedly, there is growing confidence in the integrity of human institutions. In recent years, cultural evolution theory has demonstrated the role of cultural inheritance in adapting species to social and biological environments. Moreover, scholars have emphasized unintended consequences associated with institutional change, reflecting limits on human rationality. 

These developments do not entail a case against all change – in many contexts, policy interventions remain reasonable and necessary – yet shift the basis of calculations. Compared to decades ago, conservatism now appears more reasonable.

This is not an endorsement of authoritarianism. Though certain definitions of conservatism emphasize hierarchy and discipline, I understand conservatism as the defence of established institutions. Aside from the archetypal ways in which conservatives oppose change, small-c conservatives might resist change in atypical areas, such as resistance to abortion bans, labour market deregulation and Brexit. 

One development presents a serious challenge to my argument: the increasing likelihood of exogenous shocks, primarily associated with the climate crisis. These will justify profound reforms, involving the ruptures which conservatives oppose, yet enhance other arguments for conservatism; if there are great changes in some areas, there is a stronger case for minimizing change in other domains.

2.

Ideologies are associated with epistemic regimes; in other words, different political views reflect how much we think we know about the world. For example, those with high degrees of certainty tend to be less tolerant of others, totalitarian regimes embodying this tendency. Mercifully, liberal democracy is associated with lower degrees of certainty, this system allowing dissension from governing ideology.

But even within liberal democracies, there are distinct levels of epistemic certainty. The high point might have been over a century ago, the 19^th century being an optimistic time, but recent decades have seen comparative peaks. For example, evidence-based policymaking was popular twenty years ago. This approach seemed eminently reasonable; who could be against evidence-based policymaking?

Many had practical objections to this – the third-way social democrats who favoured evidence-based policymaking tended to find evidence consistent with third-way social democracy (!) – yet few disputed that evidence could be coherent. Postmodernism had existed for decades, but was somewhat irrelevant, quantitative social scientists and policymakers dismissing it as eccentric philosophy. About ten years ago, this changed. 

Starting in psychology, scholars began to realize that they could not replicate findings. In 2015, researchers replicated 100 leading studies in the field of psychology, finding that only 36% of the replications yielded significant findings. Moreover, the mean effect size in the replications was about half the effect reported in the original studies. In other fields, similar crises followed. The sciences were not spared; in a Nature survey, over 50% of researchers admitted failing to reproduce their own experiments. 

Recent studies confirm such problems. A new paper shows that half of the effect size in observational studies, even in conservative estimates, is composed of confounding. In layperson’s terms, many studies considerably overestimate the effect of the variables which they study. Other studies show that single datasets may yield multiple findings. In a 2018 paper, 29 expert teams used a single dataset to analyze whether football referees were more likely to red card dark-skin-toned players, the teams coming to radically different results, none of which could be proved incorrect. 

Whilst these developments do not discredit academic findings – indeed, a recent paper cast doubt on the referee study – they have diminished confidence. A decade ago, a quantitative social scientist could scoff at postmodernist critiques (and many did!). The wilder claims of postmodernists remain unfeasible, yet critiques of quantitative studies are no longer easily dismissed. Depressingly, many problems reflect academic practices. Over-publication is a serious problem, encouraging the cherry-picking of interesting findings and the publication of papers which have little practical relevance.  

3.

Whilst declining epistemic confidence may seem esoteric, the importance of this development should not be underestimated. There will be direct effects; for example, expert advice to politicians may become less certain. More significantly, scepticism in academic findings might correspond with changes in wider society. Recently, there has been diminishing faith in great Western institutions such as liberal democracy and capitalism. Though causal relationships are difficult to untangle, this may be related to declining epistemic confidence; Thomas Kuhn, the famous philosopher of science, showed that societal paradigms yield diminishing returns. 

This downgrades the causal role of developments in academia, yet endows such processes with systemic significance; we may be living in an age of disbelief! If so, the importance of such events cannot be underestimated. And public attitudes are likely to reflect this, voters having less confidence in policy interventions. 

These developments have key implications for ideologies. Progressive ideologies, such as social democracy and liberalism, may fare worst. At their core, these ideologies promise that interventions can improve the existing system. Economic growth supports this vision, conditions tending to be favourable during periods of social-democratic success. This entailed a virtuous circle, culminating in the mantra of evidence-based policymaking. 

Though moderate progressivism will always appeal – it enables the hedging of bets (!) – declining epistemic confidence is a significant blow against such ideologies, undermining their raison d'être. Poor economic conditions have compounded this, growth and productivity weakening in recent decades. Contrasting with post-war decades, the combination of bad economic conditions and declining epistemic confidence entails a vicious circle.

In these conditions, radicalism becomes more appealing, this being predicated on the notion of rupture. For radicals, the existing system is rotten, entailing the need for abrupt change. Most radical intellectuals are left-wingers, emphasizing problems with capitalism, but this category includes the radical right, such movements also advocating a breach with the status quo. Admittedly, contemporary calls for rupture tend to be less ambitious than historic ones – bloody revolutions are less attractive to the inhabitants of postmaterial societies (!) – but such preferences remain clear.   

On the one hand, declining epistemic confidence scarcely gives one positive reason to advocate ruptures. Radical visions of ruptures are infamously problematic. Marxist visions of post-revolutionary periods are the weakest part of the ideology; in the 20^th century, these utopias became nightmares. 

But whilst declining epistemic confidence does not provide a positive case, it undermines the case against radical competitors. If it becomes more difficult to prove that moderate policies will improve society, the appeal of such agendas becomes reduced. Relatedly, risks associated with ruptures become more acceptable. 

Receptiveness to these positions depends upon one’s ideological priors. As moral foundations theory demonstrates, arguments tend to reflect the prejudices of authors, rather than the merits of evidence! Notwithstanding this, one ideology specializes in the negotiation of uncertainty. 

Conservatism is commonly misunderstood; many causes associated with the political right, such as deregulation and Brexit, can be considered unconservative, disrupting institutions rather than reinforcing them. I understand conservativism as the defence of established institutions. This is a minimal definition of conservatism, others emphasizing belief in order and divine intent, yet reflects the most longstanding concern of the ideology.   

Edmund Burke expressed this position most cogently. In Reflections on the Revolution in France, Burke critiqued early stages of the French Revolution, asserting that wisdom was embedded within institutions, reflecting long-term gestation. Resultingly, sudden ruptures with institutions are counterproductive, inviting disorder and tyranny. But Reflections does not discount the possibility of change, arguing that institutions should be improved gradually; Burke admired the English constitution. 

Burkean conservatism has existed for centuries, yet recent developments are consistent with its tenets. Cultural evolution theory is a seminal approach, building upon theories of biological evolution yet emphasizing the role of cultural inheritance in adapting species to social and biological environments. In other words, institutions have developed to suit human need. A major body of research supports the theory. Whilst we should not trust research blindly, as the replication crisis demonstrates (!), nor should we be completely dismissive. And cultural evolution theory has a strong evidence base, numerous studies demonstrating its logic. 

Relatedly, scholars have recognized the complexity of institutions. Traditionally, researchers evaluated institutions using rational-choice theories; such approaches interpret political institutions as the products of rational pacts, actors being aware of the consequences of actions. In recent decades, such approaches have been critiqued, scholars emphasizing the limits of rationality and the possibility of unintended consequences. This implies that interventions can have negative consequences, strengthening the case for favouring the status quo.

Whilst these are not the only relevant developments – below, I discuss the increasing potential for exogenous shocks - I think that they strengthen the case for conservatism. When the potential rewards of change decrease, the status quo becomes more appealing. This is not a case against all policy interventions! In certain cases, interventions remain unavoidable and imperative. Identification of two ideal types will enable us to establish the conditions in which interventions are advisable. 

Firstly, there are limited policy interventions. In these cases, success metrics are well-defined and potential consequences are limited. For example, the government might forbid a building material which has been demonstrated to be flammable. Whilst such a measure is associated with broader success metrics (i.e. rates of accidents at work and fire emergencies) and compliance can never be guaranteed, the policy itself fulfils an important goal, outlawing a dangerous material. 

Moreover, implications are limited; forbidding such a material is unlikely to have unintended consequences which will outweigh the benefits of the ban. Given these conditions, declining epistemic confidence has fewer implications. 

Secondly, there are expansive policy interventions. In these cases, success metrics are difficult to establish and consequences are open-ended. Such interventions can be individual, but may be part of a package. For example, a government might deregulate the labour market reform with the aim of lowering unemployment.

It has always been difficult to establish the success of expansive interventions. One may hypothesize that Reagan or Thatcher labour market reforms reduced unemployment, yet one can never conclude this decisively, at least with current methods. Establishing causality is too difficult, multiple variables interacting with each other. And these problems are separate from the recent decline in epistemic confidence, this further eroding confidence in evaluations. 

Relatedly, expansive interventions are associated with unintended consequences. Sometimes, such interventions may achieve desired ends, but others may entail unintended, negative consequences; in most cases, a mixture is likely. For example, Reagan and Thatcher reforms arguably lowered unemployment, yet were associated with the erosion of social institutions, including the families and churches which the right purport to hold dear! For such reasons, many Burkeans oppose neoliberal deregulation.

Most policies fall between these ideal types, yet this typology demonstrates relevant tensions. Policymakers must decide on a case-by-case basis, Burke emphasizing the importance of wise legislators. Drastically curtailing interventions is implausible. For example, abandoning all labour market reform would entail stagnation. But critically, the basis of calculations has changed. Years ago, there were better reasons to support policy interventions. Today, there are weaker grounds for doing so. Previous ages have been progressive, but we may now be living in a conservative epoch; in many areas, the case for doing less is stronger than it has been for decades. 

4.

For as long as conservatism has existed, critics have underlined problems such as tendencies towards staidness and difficulties achieving justice. My argument does not invalidate perennial criticisms – these can never be invalidated (!) - my point is merely that conservatism is more reasonable in current circumstances. But one contemporary problem presents a serious challenge to my argument: the increasing likelihood of exogenous shocks.

Exogenous shocks are unpredictable events that have profound effects. They are historically common, wars and plagues tormenting traditional societies, though diminished in the West as societies became more peaceful and affluent.

Future risks of exogenous shocks are chiefly associated with the climate crisis. Certain right-wingers are unconvinced (!), yet overwhelming majorities of scientists worry about rising global temperatures, many predicting catastrophes such as rising sea levels, droughts and species loss. But the climate is not the only source of potential instability, nuclear proliferation and risks associated with artificial intelligence (AI) also being concerning. Other crises may arrive unheralded, à la Coronavirus. 

These shocks will necessitate policy interventions. Certain measures might be incremental, as conservatives prefer, but others will be sudden, reflecting the unpredictable nature of such shocks. The Coronavirus crisis involved deep ruptures with traditional living and working practices, albeit temporary ones. The desirability of future ruptures will be contested, reflecting ideological sympathies, but the occurrence of such ruptures seems probable. 

Expansive interventions will have unintended consequences, Coronavirus demonstrating this. Though restrictions have been lifted, life has profoundly changed, people embracing remote working and moving from urban areas. Reforms associated with the climate crisis may be more expansive, revolutionizing social relations. Traditional conservatives will dislike much of this, certain liberties being threatened and societies being transformed in an unpredictable manner. 

Such measures may be more consistent with the radical tradition, this being predicated on the notion of rupture. On the environment, left-wing radicals have strong arguments, this crisis reflecting capitalist patterns of accumulation and expansion. Radicals will scarcely wish to stop with such measures, proposing ruptures in other policy areas.

Yet if one advocates changes in certain areas but worries about overshoot, conservatism offers a way forward; for if expansive measures are introduced in some domains, involving dislocation, there may be a stronger case for limiting reforms in other spheres. As conservatives grasp, change is unsettling and disruptive; stable conditions in other domains will reassure those who fear change.

For example, if climate change necessitates large-scale population transfer, such conditions scarcely encourage destabilizing labour market reform or large immigration quotas! During the Coronavirus crisis, this principle was important. Policymakers maintained focus on the pandemic, associated reforms attempting to minimize further disruption; income-replacement schemes sought to stabilize conditions in labour markets. Most accepted the validity of this. 

Indeed, high incidences of exogenous shocks are traditionally associated with conservative values. In traditional societies, conservative values provide reassurance. Whilst future shocks will (hopefully!) not reproduce the uncertainty of traditional societies, there is a clear parallel; during times in which certain changes are pronounced, policymakers might limit changes in other areas.

5. 

Analysis of the crisis of epistemic confidence encounters a paradox; if there is less reason to believe academic findings, there may be less reason to trust accounts which advance prescriptions! And as I have noted, there is increasing recognition that arguments reflect the prejudices of authors, rather than the merits of evidence. Conservatives have a peculiar profile, tending to be more risk-averse and anxious; fear of death predicts conservative values!

Relatedly, my argument is difficult to falsify. Even if one assumes that hypotheses can be falsified, problematic in an age of declining confidence, my argument is broad, being composed of multiple hypotheses; the accumulation of satisfactory evidence would take years.  

These are generic problems which should not preclude broader arguments – if essayists had to clear such hurdles, we would be in a bleak place (!) - but limit the potential of my argument. Ultimately, this essay reflects my prejudices and requires a wider evidential base.

Yet this does not mean that all arguments are equally valid. Certain arguments are more reasonable than others, notwithstanding their tentative status; societal consensus is one source of external validation. Conservatism may not be the only potential response to declining epistemic confidence, but I think that it is a reasonable one. In the face of uncertainty, many dislike turning the wheel. 

If you enjoyed this, please think about subscribing to my Substack (https://thomasprosser.substack.com)

11 comments

Comments sorted by top scores.