Book notes: Tyler Cowen, Stubborn Attachments

Tyler Cowen’s Stubborn Attachments was published in book form by the lovely Stripe Press in late 2018. However, he had published essentially the same content in a Medium post back in 2017 which has since been taken down. It is a reasonably quick, enjoyable read. I had read the Medium post and have since read the book as well. Some day I might right down my notes on the book, but below are my notes from the Medium post instead. It reads like an economist’s attempt at writing a “how to live” philosophy book reminiscent of ancient Greece or some Vedic texts.

Conclusion / punchline

• Believing in the overriding importance of sustained economic growth is more than philosophically tenable and it may be philosophically imperative. We should pursue large rather than small benefits and we should have a deep concern for the more distant future, rather than discounting it exponentially.
• We should subject ourselves to the constraint of respecting human rights, noting that only semi-absolute human rights will be strong enough to place any constraint on pursuing the benefits of a higher rate of sustainable economic growth.
• We should be very cautious in our attitudes about specific policies. Even if we succeed in taking true aim at what we think are the best courses of action, the chance that we are right on the specifics - even if as high as possible - still is not very high. Most likely you’re wrong - even if some others are likely to be even more wrong - and thus your attitudes should be correspondingly modest in the epistemic sense.
• Concrete recommendations:
  • Policy should be more forward-looking and more concerned about the more distant future.
  • We should place a much higher priority on investment than is currently the case.
  • Policy should be more concerned with economic growth, properly specified, and policy discussion should pay less heed to other values. No exceptions, except of course for the semi-absolute human rights.
  • We should be more concerned with the fragility of our civilization.
  • At the margin we should be more charitable but we are not obliged to give away all of our wealth. We do have obligations to work hard, save, invest, and fulfil our human potential, and we should take these obligations very seriously.
  • We can embrace much of common sense morality, while knowing it is not inconsistent with a deeper ethical theory. Common sense morality also can be reconciled with many of the normative recommendations which fall out of a more impersonal and consequentialist framework.
  • When it comes to most “small” policies, affecting the present and near-present only, we should be agnostic because we cannot overcome aggregation problems to render a defensible judgment. The main exceptions here are the small number of policies which benefit virtually everybody.
• We should strive for significantly augmenting future human well-being as experienced over long spans of time. As we struggle for great achievements, we will encounter many incidental costs and obstacles along the way. We are not certain how much weight we should attach to these costs given the high generic uncertainty associated with our choices. But these incidental costs - and benefits - are less important than we had thought. They should be less of an obstacle to our attempts to improve our civilization and render it more durable.

In more detail:

Starting points

• We need a tougher, more dedicated, and indeed a more stubborn attachment to prosperity and freedom than is currently the case.
• Underlying philosophical stance
  • Right and wrong are very real concepts which should possess great force. Relativism is a non-starter and most people are not sincere in their relativist pronouncements in any case.
  • We should be sceptical about the powers of the individual human mind.
  • Pluralism is considered a core moral intuition. Human life is complex and it offers many different goods, not just one trump value.
• Tyler Cowen calls himself a “two-thirds utilitarian” in that he first looks to human well-being when analysing policy choices. That said, human well-being is not an absolute priority. Sometimes, we ought to do that which is truly just, even if it is painful for many people involved.

Six critical issues

• Time: future should not be discounted as much is commonly the case (where the future declines in moral importance as time passes).
• Aggregation: how should we resolve disagreements and how do we decide that the wishes of one individual should take precedence over the wishes of another?
• Rules: do rules hold independent force in our moral reckoning? Can we generate a coherent morality in which we should respect rules and principles for their own sake? Yes.
• Radical uncertainty: Tyler Cowen is a sceptic but one with a “can do” temperament and who realizes how paralysing scepticism can be. Given long-run uncertainty, how can we pretend to assess good and bad consequences as a product of our actions? How can we make any decision at all without being morally paralysed and totally uncertain about what we are doing?
• Rights: how is it that we can believe in rights? Tyler Cowen does believe in (nearly) absolute human rights.
• Common sense morality: that we should work hard, take care of our families, and live virtuous but self-centred lives, while giving to charity at the margins and helping out others on a periodic basis. Utilitarian philosophy, on the other hand, appears to suggest an extreme degree of self-sacrifice.

On wealth and growth

• Economist Frank Knight wrote of the Crusonia plant - a mythical, automatically growing crop which creates more output each period, i.e., a free lunch. For example, an apple tree produces apples, whose seeds produce more apple trees, etc. - a compounding feedback loop. Of course, in reality, you need sun, rain, etc. as natural resources for each apple tree so it’s not a free lunch.
• When it comes to making tough decisions, we should try to identify which elements in the choice set resemble a Crusonia plant, i.e., ongoing and self-sustaining surges in value.
• Look for social processes which are ongoing, self-sustaining and which create rising value over time. The natural candidate for such a process is economic growth or some modified version of that concept. Standard definitions of economic growth do not fully qualify as true Crusonia plants, in part because they ignore environmental sustainability and they do not adequately value leisure time.
• Use the standard of “Wealth Plus”: accumulated gains from growth which accounts for leisure time, household production (valuable activities you do at home for free, whether mending socks or using Facebook), and environmental amenities, among other adjustments.
• Three key questions should be elevated in their political and also philosophical importance, namely:
  • What can we do to boost the rate of economic growth?
  • What can we do to make our civilization more stable?
  • How should we best deal with environmental problems?
• Interestingly, the first is generally libertarian; the second is generally ‘conservative’, and the third is ‘left wing’.
• Tyler Cowen roughly seems to be supportive of trickle down economics (in this case, the growth of wealthier countries trickling down to others). Although these historical processes have often embodied unfairness and long lags, of decades or more, economic growth nonetheless has brought wealth to the poor and elevated their status.
• Tyler Cowen argues that economic growth makes us happier.

Overcoming disagreement

• In order to resolve disagreements (whose wishes take precedence?), use the principle of growth: We should make political choices so as to maximize the rate of sustainable economic growth, as defined by Wealth Plus.
• Or rather, the modified principle of growth: We should push for sustainable economic growth, but not at the expense of inviolable human rights.

Exceptions to the rules?

• No, not really.
• Parfit’s Mistakes in Moral Arithmetic: If a firing squad of six shooters kills an innocent person, can we say that any one of the shooters is a murderer? After all, the ‘marginal product’ of any single shooter was zero. Should we punish or invest resources to prevent any one of the shooters? Does it matter whose bullet arrived first? Should we refuse to prosecute group murders of this kind?
• Evidently, participating in such a shooting is wrong.
• While the example is stylized, it fits many real world situations. Maybe a single act of corruption has no harmful effects but corruption in general is harmful and many corrupt acts will destroy a polity.

Time, or discount the future less than you do

• Cowen and Parfit (1992) argues that the social discount rate as commonly applied is too high. For example, if we discount the future by 5%, a given death today is worth about 39 billion deaths 500 years from now.
• At the very least, we should be sceptical that positive discount rates apply to every choice before us.
• Even if you think that individual impatience is (sometimes) justified, it will not justify the positive discounting of well-being across the generations. It cannot be argued that medieval peasants benefited from having been born before us and thus having eaten their bread sooner. Time preference therefore does not justify the significant discounting of the distant future, even if it justified Tom wanting a steak dinner sooner rather than later.
• That said, discounting for risk is justified in a way that discounting for the pure passage of time is not justified. But if we boost the long-term sustainable growth rate, we are making the future less risky.
• Caveat: the opportunity cost argument. Investing a dollar today, at positive returns, will yield more than a dollar next year.
  • These arguments deal with wealth only and they do not establish positive discounting across well-being. Well-being is an end in itself but wealth is a conduit to well-being and other plural values.

Redistribution, or do we need to give away most of our wealth à la Singer’s recommendations?

• While we are obliged to help the poor more than we are doing now, the correct approach does not lead to personal enslavement or massive redistribution of our personal wealth.
• Our strongest obligation is to contribute to sustainable economic growth and to support the general spread of civilization, rather than to engage in massive charitable redistribution in the narrower sense. We should redistribute only up to the point which maximizes the rate of sustainable economic growth.
• We should reject collective sacrificial recommendations that will lower the rate of sustainable economic growth.
• Consider a game with pay-offs such that it is better if someone sacrifices, to achieve a socially valuable end, but it is worse if everyone sacrifices or tries to sacrifice to achieve that end.
  • The structure of this problem is common to many questions of morality and individual obligation, including the problem of global poverty. Some people should make sacrifices to help out but, because we must keep an economically advanced civilization up and running, not everyone should make such sacrifices.
  • The observed response (that the “someone” is going to be someone else) is not socially optimal, but that the best solution to the game corresponds to some underlying features of common sense morality.
  • Recommendation: we should strengthen our consciences, and strengthen social norms, to increase the probability that the appropriate individuals would be willing to make a needed sacrifice, if it turned out to be best, all things considered, that they be the ones who should step forward. We ought to honour and reward such sacrifices more, to increase their likelihood, and again this is not so far from common-sense morality.
• Should redistribution from poor to rich occur?
  • For example, if there is a trickle-down effect from the wealth of the wealthy, combined with a zero rate of discount, it is easy to generate scenarios where utilitarianism recommends redistribution to the wealthy.
  • Usually this trickle down won’t come right away, but over time. Sooner or later a lot of the poor will benefit. If we have a deep concern for the distant future (i.e., nearly zero discount rate), it matters less if a lot of these benefits come later.

Uncertainty, or should it paralyse us?

• The epistemic problem, that we cannot know the effects of our current actions.
• At some point we can find a set of consequences so significant that we would be spurred to action, without much epistemic reluctance, even though we would be recognizing the broader uncertainties of the very long run.
• Once the upfront benefit from one course of action is sufficiently high, the epistemic critique has less force, though we remain uncertain as to whether we will choose the correct beach for defeating Hitler.
• The Principle of Roughness: “Some of our choice options will differ in complex ways. We might nonetheless, ex ante, make a reasoned judgement that they are roughly equal in value, and that we should be roughly indifferent across the two options. After making a small improvement to one of these choices, we still might be roughly indifferent to which option is better.”
• If there is any victim to the epistemic critique, it is focusing on small benefits and costs, but not consequentialism more generally. If we bundle appropriately and “think big” and pursue Crusonia plants our moral intuitions will rise above the froth of long-run variance.

How to be most likely correct

• How to be a good agnostic
  • Given the radical uncertainty of the more distant future, we do not usually have a good idea how to achieve our preferred goals over longer time horizons. Our attachment to particular means therefore should be highly tentative, highly uncertain, and radically contingent.
• Why the case for rights is compelling and which rights are the important ones?
  • In other words, very often rights do not conflict with consequences in the simple ways set up by philosophic thought experiments. And so now we can see a shift in how we can think about radical uncertainty concerning consequences. Rather than letting it paralyse us, think of the uncertainty as giving us the freedom to act morally, without fear of the certain knowledge that we are engaging in consequentialist destruction.
  • Murdering one baby or aliens from Alpha Centauri destry the entire earth: It is possible to see why we probably would opt to kill the baby in such a case. For one thing, the cost of not murdering the baby is now much higher. For another - and this is significant - if the entire world ends there is no residual uncertainty about what will happen next. Arguably we should pursue the better consequences and there is no remaining “froth of uncertainty” to justify sticking with the rights of the individual baby. And in that sense the notion of rights postulated is not strictly absolute against all possible external consequences, as might be dreamt up by philosophers. Still, most or all of the hypothetical examples where rights should be violated are not very relevant to the real world choices we have had to face so far.