by Tyler Cowen
Second, even putting well-being aside, the whole point of the opportunity cost argument is that material wealth today can be invested and earn a positive rate of return, at least on average. That’s a good argument for substantial investment in the future; it is not an argument for caring less about the future.
Another argument for strong discounting of the future is that those individuals will be better off or wealthier than we are today. But even if we take the egalitarian premise of this argument for granted, it will lead to discounting for wealth rather than across time. The discounting would apply to today’s wealthy people, too, but our deep concern for the distant future will continue to hold. For one thing, a lot of individuals, even in the future, probably won’t be wealthy at all, or at least they won’t be wealthy unless we do the right thing. In essence we are choosing their wealth through current policies (if we choose properly), so we should not assume their wealth as a justification for neglecting them. Or maybe they’re wealthy for most of their lives, but they’re not very wealthy when faced with extreme tragedy or serious illness. No matter how bright the future may seem, it’s still going to yield an ongoing stream of human tragedies. The way to minimize those tragedies, again, is to maximize sustainable growth.
Rather than opting for a strictly zero discount rate, I suggest a more modest postulate, one to which I already have referred but will now label formally: Deep Concern for the Distant Future. In this view, we should not count catastrophic losses for much less simply because those losses are temporally distant. In the absence of qualifying factors, no amount of temporal distance per se should cause major widespread tragedies to dwindle into insignificance in the present. We should believe that the end of the world is a truly terrible event, even if that collapse comes in the distant future. Similarly, the continued persistence of civilization three hundred years from now is much better than having no further civilization at that point in time. A much wealthier future civilization is much better than a less wealthy future civilization. Those are the implications of a deep concern for the distant future.
If you are technically and mathematically minded, you can later read Appendix A, in which I lay out a version of what is called the overtaking criterion. That is one formal way to pin down a deep concern for the distant future, but you don’t need mathematics to grasp the intuition. The overtaking criterion implies the following: given the long-run comovement of sustainable growth and human well-being, if one growth path is consistently higher than the other over time, we should prefer that higher growth path. At some point in the future, the higher growth rate will make people much, much better off, and that path is worth choosing even if it involves some cost in terms of foregone consumption today.
The overtaking criterion pushes us toward maximizing Wealth Plus, as defined in the previous chapter, yet without committing to the idea that future well-being is worth exactly as much as current well-being. So the overtaking criterion is not a final or definitive moral view on how to compare present and future values in each and every circumstance. Rather it represents one practical rule, a kind of minimum commitment to the future of human well-being. I’ll discuss this more in Appendix A, but for now I recommend simply pocketing the intuition and moving on to the next chapter.
1. On that kind of discounting, see Frederick, Loewenstein, and O’Donoghue (2002).
2. Castillo et al. (2011). See also the recent Moffitt et al. (2011) and Mischel, Shoda, and Rodriguez (1989). On the connection between patience and measures of cognitive ability, see Frederick (2005). Ipcher and Zarghamee (2011) consider the connection between positive affect and low time preference.
3. Cowen and Parfit (1992, 145).
4. To see how this works, consider a much wealthier future one hundred years from now. In spite of higher overall wealth, some individuals might die from current decisions, such as environmental neglect. If those individuals could bid on today’s policies, they would bid for greater concern for the future. If we think of the growth of wealth as (roughly) matching the interest or discount rate, their bids, in today’s terms, would carry about as much weight as ours. And if the policy had the potential to make a big impact on their lives, future generations could well be the dominant bidders in a policy evaluation. Yet current market prices are not giving them any weight, or are giving them weight only through our imperfect altruism. Think of this as further evidence that current methods are undervaluing the allocation of resources to future generations.
5. On how individuals respond to queries about current and future lives, see Frederick (2003, 2006).
6. Note, by the way, that individual time preference is a tricky concept. What if I am very hungry and wish to eat now rather than later? Does this count as time preference? Or does time preference imply that the same “eating experience”—holding the level of hunger constant—should be preferred sooner rather than later? If we hold everything constant, in what sense can we even say that time has passed? So the passage of time means that ceteris paribus clauses cannot be totally strict, but then we return to the question of what should vary with the postulated passage of time. Arguably all instances of “goods in different time periods” are simply different goods, noting that economists define “goods” in terms of revealed preference and indifference. These messy problems haven’t been cleaned up, but they do cast further doubt on the idea of the rationality of purely positive time preference per se.
7. Spolaore and Wacziarg (2013) survey this research. Acemoglu, Johnson, and Robinson (2001, 2002, 2005) show how decisions and institutions from the early colonial era still affect economic performance today. Comin, Easterly, and Gong (2010) show the extent to which regional wealth in 1000 B.C. still predicts regional wealth today.
8. Cowen (1997) pursues these questions.
9. See Lind et al. (1982), Arrow et al. (1994), Broome (1994), Brennan (2007), and Gollier (2013). Cowen (2007) also discusses these issues.
10. This argument for discounting is more complex than is sometimes recognized. It does not require time preference, but it does require assumptions about the intertemporal substitutability of consumption. Diminishing marginal utility, in the classic sense, is defined at a single point in time. But how do differing marginal utilities of consumption vary across time? How does my two millionth dollar next year compare to my one millionth dollar today? This variable is distinct from either classic time preference or classic diminishing marginal utility. For the classic economic argument to work, it is usually assumed that consumption tomorrow is a relatively close substitute for consumption today.
5—What about redistribution?
Why eat an ice cream cone when someone in Malawi is starving? That’s one version of an age-old question in ethics, yet it remains difficult to answer.
Under any moral theory which counts the interests of people in a more or less cosmopolitan manner, our personal obligations toward the poor appear strong. For instance, several billion people in the world currently live on less than two dollars a day. Last year millions of children died of preventable diseases such as diarrhea or experienced stunted growth and development. Presented with these examples, it’s easy to feel like we should all be attending to such problems with more resources and more energy than we do now, and indeed we should. The more difficult question, however, is how far such obligations extend and whether such obligations should prevent us from pursuing our own more personal or more individualistic goals. In contrast to utilitarianism, common sense morality typically suggests that we do have the right to pursue our own lives and life plans.
By some accounts of our obligations to others, every individual is obliged to work a certain number of years for charity, or to send most of his or her income to the poor. Wealthy doctors should spend large parts of their careers, if not their entire careers, in African villages. Many more of us would have to become doctors or nurses, unless of course you would be more effective as a Wall Street wiza
rd and then a wealthy philanthropist. Maybe these options sound meritorious to you, but how far are you willing to go to act on your beliefs? Are you willing to value the interests of others on par with your own, or those of your family and friends? If you buy into the standard utilitarian logic of beneficence, a mother might have to abandon or sell her baby in order to raise money to send food to the babies of others. At this point most people balk at the argument and search for some moral principle that limits our obligations to the very poor.
One problem is that the needs of the suffering are so enormous that only a few able or wealthy individuals would be able to carry out individual life projects of their own choosing. Most people would instead become a kind of utility slave, serving only the interests of others and feeding themselves just enough to survive. The result is that utilitarianism—or many forms of consequentialism, for that matter—is often seen as an excessively demanding moral philosophy. People fall into two camps: those who reject utilitarianism for its extreme and unacceptable implications, and those, like the early Peter Singer, who trumpet the call for greater sacrifice and pursue the utilitarian logic to a consistent extreme.
As I stated earlier, I am a pluralist rather than a simple utilitarian. Still, utility is a central and important value, so I would like to confront these dilemmas and consider the scope of our obligations to the poor. I don’t, however, wish to focus on government vs. private charity. To be sure, that is an important practical issue, but I’ll instead focus on the broader conceptual question of whether growth or redistribution—in the public or private sector—is a more effective means of helping the poor. When framed in this manner, we’ll see that there are some strong and strict limits on our obligations to redistribute wealth, even if we accept a full utilitarian framework. In other words, I’ll end up siding with common sense morality, which (mostly) allows us to pursue individual life goals.
It can be argued very plausibly and, I think, correctly that we are obliged to help the poor more than we are doing now. But the correct approach to our cosmopolitan obligations does not lead to personal enslavement or massive redistribution of our personal wealth. Most of us should work hard, be creative, be loyal to our civilization, build healthy institutions, save for the future, contribute to an atmosphere of social trust, be critical when necessary, and love our families. Our strongest obligations are 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. In the longer run, greater economic growth and a more stable civilization will help the poor most of all.
This point can be expressed as follows: we should redistribute wealth only up to the point that it maximizes the rate of sustainable economic growth. This may mean more redistribution than we currently undertake, and sometimes redistribution of a different kind, namely growth-enhancing redistribution. (By no means do all of today’s government programs actually redistribute wealth to the poor, much less boost economic growth.) It will not, however, suggest that a utilitarian or consequentialist approach obliges us to redistribute most of our nation’s income to the very poor. Nor are productive individuals morally required to spend most of their years serving the very poor.
There nonetheless remains a good case to be made for some degree of redistribution. For instance, a well-constructed welfare state can distribute education and nutrition more widely. The individuals supported by this state are not only better off, but they are more likely to be productive and pay taxes, and they are less likely to overturn public order. Other benefits of redistribution stem from political improvements. Social welfare programs can buy the loyalties of special interest groups and decrease feelings of desperation, both of which can help cement social order. Furthermore, social welfare programs make many citizens feel better about their state, again boosting public order as well as political consensus and stability.1
These factors rationalize some investments in a welfare state—and yes, we should describe them as investments—and thus support some wealth redistribution. Furthermore, they suggest an appropriate nature and scope for such redistribution, namely that we try to enhance sustainable economic growth.
Past some margin, an overly generous level of wealth transfer harms economic growth. Many people end up working less, or working less hard, and the associated higher tax rates discourage entrepreneurship and can lead to economic stasis. Furthermore, if it is standard procedure to approach government for a handout, that will induce too much rent-seeking, dependency, corruption, and eventually fiscal imbalances and perhaps even insolvency or a financial crisis. Alternatively, excess or poorly conceived welfare expenditures may create urban cultures of dependency and crime, which endanger social order. The empirical literature suggests that non-infrastructure government spending is correlated positively with lower growth rates, with the caveat that those results are measuring traditional GDP rather than an alternative notion such as Wealth Plus.2
Excess transfers are bad for another reason, namely that they make it harder to absorb high numbers of immigrants from poorer countries. Whether rationally or not, many voters feel that migrants are a burden on the government budget, and they resent it when some immigrants receive government benefits. To be sure, some level of social welfare makes immigration run more smoothly by helping immigrants to become established and self-sufficient. But too high a level of benefits is likely to mean, one way or another, a lower level of migration and a corresponding reduction in the most effective anti-poverty program we have discovered to date. This too should limit our attachment to social welfare programs.3
So rather than redistributing most wealth, we can do better for the world by investing in high-return activities like supporting immigration and producing new technologies with global reach, such as cell phones and new methods for boosting agricultural productivity. Many people mock the term “trickle-down economics,” but most social benefits do take a trickle-down form. We should of course prefer a flood to a trickle, which brings us back to wishing to boost the sustainable growth rate as much as possible.
These stipulated individual obligations are not so far from common sense morality. To be sure, we have not yet bridged the gap between utilitarian reasoning and common sense morality. Even when utilitarianism and common sense recommend the same course of action, they do so for different reasons. Utilitarianism tells us we should work, save, and innovate to serve the purposes of others, including future generations. Common sense morality tells us we should work and save to take care of our families and because we value our own lives. These two perspectives remain distinct in their methods and their justifications. Nonetheless, to the extent that the practical conclusions converge, we can think of utilitarian and common sense modes of reasoning as two aspects of some broader moral picture. After all, I favor pluralism rather than utilitarianism or common sense morality per se. So we do not have to bring these two perspectives into complete accord. Instead, for the purpose of constructing a workable pluralism, it may suffice to know that two of the tools in our toolbox point in broadly compatible directions.4
We can now consider the question of why public and private ethical codes might differ so much in their recommendations. In the private sphere, virtually everyone would agree that a mother is justified in looking after her own baby rather than selling it to send resources to the babies of strangers. Yet in the public sphere, it is widely believed that governments should be more impartial, at least within national boundaries.5 Governments should not favor the interests of one particular baby over another, at least not for the same reasons that a mother would favor her baby over a stranger’s.
But why should morality be so bifurcated? Simply reiterating the existence of our intuitions won’t justify that practice. But within the framework of this book, perhaps we can see why moral obligations vary between an individual acting in a private capacity and a government acting in the interest of increasing the
rate of sustainable economic growth. This is actually just another application of Adam Smith’s classic principle of the division of labor. It’s easy to see why parents do best when they attend to their family priorities and governments do best when they institute a legal regime based on impartiality and the rule of law. Again, I’m not suggesting that this argument provides an exact justification for the intricacies of common sense morality as it is appears in the world, with all of its quirks and custom-dependent variations. I’m simply noting that the gap between common sense morality and utility maximization, properly construed, is much smaller than it might at first appear.
We can test these hypotheses by looking for cases in which a utilitarian or a consequentialist should favor large-scale redistribution toward the very poor. For instance, the case for redistribution would be stronger if the world were going to end in the near future. If the time horizon is extremely short, the benefits of continued higher growth will be choked off and the scope for compounding over time would be correspondingly limited. The immediate returns to charity should therefore weigh more heavily in the decision calculus. To present this point in its starkest form, imagine that the world were set to end tomorrow. There would be little point in maximizing the growth rate; arguably, we should just throw a party and consume what we can (after feeding the hungry, that is).
Alternatively, the real return on investment might be permanently negative or zero. In this case, compounding of returns would not operate, a long time horizon would not lead to a Crusonia plant, and there would again be greater incentive to redistribute wealth. A high degree of redistribution also makes sense in a lot of “lifeboat” settings, where, for example, a group of desperate individuals are afloat on the open sea and are wondering how to share the cans of Spam. These examples typically involve an implicit assumption of a zero or negative rate of return on investment, or perhaps investment simply isn’t possible in such an artificial setting. Again, there isn’t a Crusonia plant to chase after.