Stubborn Attachments

by Tyler Cowen

  Very often the choice between the present and the future takes place at the social level. Many social policies influence whether benefits and costs come sooner or later, and if we are to make a choice, we need to decide how impatient we are going to be. I worry about the logical implications of impatience, if we were to apply such impatience to a longer time horizon. Together with Derek Parfit, I once wrote:3

  Why should costs and benefits receive less weight, simply because they are further in the future? When the future comes, these benefits and costs will be no less real. Imagine finding out that you, having just reached your twenty-first birthday, must soon die of cancer because one evening Cleopatra wanted an extra helping of dessert. How could this be justified?

  Economists and other social scientists often speak of a “discount rate.” A discount rate tells us how to compare future benefits to current benefits (or costs) when we make decisions. When the discount rate is high, we are counting future costs and benefits for less. Let’s speak in terms of pleasure (or pain) as a magnitude that corresponds, however roughly, to a real number scale. A five percent discount rate, defined annually, means that 100 units worth of pleasure today is equal to 105 units worth of pleasure a year from now. A ten percent discount rate would set this equality at 110 units worth of pleasure a year from now, and so on.

  A discount rate of zero means that a future benefit (or cost) counts for as much as a comparable benefit in the present. A person with a zero discount rate would not see any point in putting off going to the dentist. There’s no reason not to get it over with.

  If there’s one thing we’ve learned, it’s that discount rates matter. In your personal life it affects how hard you work, how much you drink and gamble, and what kind of education you get. At the social level, the discount rate pertains to questions of how hard we should be fighting climate change and how much we should invest in preserving biodiversity. If we dismiss the importance of the distant future, action will not seem imperative. But if we pay heed to the distant future, we will see these as major concerns.

  Discounting also matters for how hell-bent we are on pursuing a higher rate of economic growth. A higher growth rate means that the future, at some point in time, will be much richer than it would be otherwise, and, as I argued earlier, it also means that human beings will be much better off. How compelled should we feel to bring about this wealthier state of affairs? If you only care about today, you won’t be as motivated to act in favor of higher sustainable growth.

  Most of us are altruistic, especially toward our own children and grandchildren. But this form of partial altruism does not make us care much about other people’s grandkids. When people vote or otherwise make choices that affect future generations as a whole, they often behave quite selfishly. Political time horizons tend to be very short, often extending no further than the next election or the next media cycle. Voters are keen to receive more government spending now and postpone the required taxes to the more distant future. Few governments do everything they can to promote economic growth for the more distant future. The bottom line is that caring about the future is not something that happens automatically, even if you dearly love, or will dearly love, your grandchildren. When it comes to the discount rate for social decisions, we need to choose wisely.

  For certain decisions, such as whether or not to cut down a tree, market forces induce even selfish people to think about the more distant future. If you leave the tree standing, it might be worth more money. If you own a Rembrandt painting, you’ll probably keep it in decent shape, even if you’re a selfish, uncultured bastard who doesn’t care about the artistic patrimony of the Dutch. These kinds of examples, however, apply only when there are well-defined property rights to specific assets. The motivations behind these behaviors won’t spur us to preserve the environment or maximize the rate of sustainable economic growth. Once again, the proper depth of concern for the more distant future does not come to us automatically, at least not in a wide variety of cases.

  Expressed differently, when it comes to non-tradable and storable assets, markets do not reflect the preferences of currently unborn individuals. The branch of economics known as welfare economics holds up perfect markets as a normative ideal, yet future generations cannot contract in today’s markets. If we were to imagine future generations engaging in such contracting, current decisions might run more in their favor. Circa 2018, the future people of 2068 can’t express their preferences across a lot of the choices we are making today, such as how rapidly to boost future wealth or how much to mitigate the risk of serious catastrophes.4

  Let’s now consider some basic choices about how to value the distant future. Again, think of a decision-maker weighing present and future interests, in this case human lives. The way discounting works, if we discount the future by five percent, a person’s death today is worth about thirty-nine billion deaths five hundred years from now. Alternatively, at that same discount rate, one death two hundred years from now is equal in value to 131.5 deaths three hundred years from now. Upon reflection, few people, putting aside their selfish interest in the current time period, would share these conclusions as a basis for ethical decision-making.5

  Or consider the comparison prospectively. Under any positive discount rate, no matter how low, one life today could be worth more than one million lives in the future. It could even be worth the entire subsequent survival of the human race, if we use a long enough time horizon for the comparison. At the very least, we should be skeptical that positive discount rates apply to every choice before us. Sometimes we should be less impatient and pay the future greater heed.

  Even if you think that individual impatience is sometimes justified, impatience will not justify the positive discounting of well-being across generations. Time preference may mean that an individual prefers to have a good steak dinner sooner rather than later. Even if this is rational—after all, you’re getting hungrier by the minute—this kind of time preference doesn’t apply across longer time frames, including future generations. Our still-unborn great-great-grandchildren will not receive benefits for some time. But in the meantime they are not sitting around, waiting impatiently with rumbling stomachs. It cannot be argued that their forthcoming slice of time is worth less simply because they must wait for it. Similarly, it cannot be argued that Medieval peasants benefited from having been born before us and thus having eaten their bread sooner. When we consider long periods of time and count the years before individuals are born, we need to discard impatience as a factor of relevance because it just doesn’t apply. Time preference therefore does not justify the significant discounting of the distant future, even if it justifies Tom’s wanting to have his steak dinner sooner rather than later.6

  Another way of thinking about why a high time discount rate is wrong involves a somewhat unusual—some would say kooky—thought experiment. Einstein’s theory of relativity suggests that there is no one factual answer to the question, “What time is it?” Any measurement of time (when is “now”?) is relative to the perspective of an observer, and to the velocity of that observer relative to the speed of light. In other words, if you are traveling very fast, you are moving into the future at an especially rapid rate. Yet it seems odd, to say the least, to discount the well-being of people as their velocity increases. If, for instance, we sent off a spacecraft at nearly the velocity of light, the astronauts would return to Earth, hardly aged, many millions of years hence. Should we pay less attention to the safety of our spacecraft, and thus to the welfare of our astronauts, the faster those vehicles go? Should we—as a result of positive discounting—not give them enough fuel to make a safe landing? And if you decline to condemn these brave astronauts to death, how are they different from other residents of the distant future?

  Instead of letting our speedy astronauts die, we can think of the universe as a block of four-dimensional space-time. We would not discount human well-being for temporal distance per se any
more than we would discount well-being for spatial location per se. In moral terms, maybe time really is an illusion, as Buddha suggested thousands of years ago.

  That said, discounting for risk is justified in a way that discounting for the pure passage of time is not. If a future benefit is uncertain, we should discount that benefit accordingly because it may not arrive. But such a practice does not dent a deep concern for the distant future. It is precisely because we discount for risk that we seek to protect our future against great tragedies, thereby making that future less risky. If we boost the long-term sustainable growth rate, for instance, we are indeed making the future less risky. Rather than ignoring risk, a future-oriented perspective takes long-term risk into account and attempts to lower it. The factor of risk might encourage you to spend your money now, otherwise someone might steal it. But it won’t discourage us from caring a lot about long-term sustainable growth.

  Before moving on, let’s consider the relevance of the numerical comparisons presented above of events which lie one hundred, two hundred, or even five hundred years into the future. It might seem that nothing we do today can affect the world that far out, most of all when it comes to policy issues. Yet the most recent evidence suggests that good (or bad) political and economic decisions, and the general existence of prosperity, have persistent effects that stretch for centuries into the future. Colonial policies from the sixteenth and seventeenth centuries have persistent effects on prosperity today, and there is even research suggesting that the prosperity of a region well before the birth of Christ holds predictive power for the prosperity of those regions today.7

  For whatever reason, good institutions and a history of prosperity tend to have enduring effects. Wealth can fund and enable better government, and that in turn gives rise to further wealth and better institutions. Institutional memories of economic success and good governance can persist for long periods of time. Cultural practices such as business savvy or an interest in external markets can last for centuries.

  England, which led the Industrial Revolution, had positive institutional features stretching far back in its history, such as relatively free labor markets in Medieval times and the carving out of a coherent national unit with a language, an army, and a parliament. The practices of the empire then carried some of these institutions across the oceans, such as when the British settled much of North America and the Antipodes (though not every region benefited from the brighter side of British rule). It’s no accident that many of the original territories of the Roman Empire remain some of the world’s wealthiest and most successful nations. China was also a relatively wealthy nation in earlier times, and that prosperity is reemerging today. For centuries, Chinese entrepreneurs around the world have shown special commercial savvy; this again has something to do with history.

  Of course, the persistence of prosperity does not apply in every case. Much of the Arab world is currently well below its historic relative standing; Baghdad might have been one of the best and most interesting cities to live in about a thousand years ago, but today it is struggling. Still, if we think in terms of averages, we see plenty of evidence that history can matter over very long time spans. Therefore, any act which strengthens good institutions today has, in expected value terms, a causal stretch running centuries into the future. Once again, this means that our choice of discount rate is of critical importance.

  We can also see the importance of faith to the overall argument. To fully grasp the import of doing the right thing, and the importance of creating wealth and strengthening institutions, we must look very deeply into the distant future. As I have argued at length, this is a conclusion suggested by reason. But in the real world of actual human motivations, the application of abstract reason across such long time horizons is both rare and unhelpful when it comes to getting people to do the right thing. The actual attitudes required to induce an acceptance of such long time horizons are, in psychological terms, much closer to a kind of faith. We cannot see these very distant expected gains, but we must believe in them nonetheless, and we must hold those beliefs near and dear to our hearts. In this sense, we should strongly reject the modern secular tendency to claim that a good politics can or should be devoid of faith.

  There are, of course, many bad forms of faith in politics, and we should not encourage political (or other) beliefs in willful disregard of reason. But we cannot kick away faith itself as a motivational tool, as politics is of necessity built on some kind of faith. The lack—and, indeed, the sometimes conscious rejection—of the notion of faith, as is common in secular rationalism, is one of the most troubling features of the contemporary world. It has brought us some very real gains in terms of personal freedom, but it also threatens to diminish our ability to make the very best choices.

  Should we discount the past?

  So far we’ve been considering how much the future is worth compared to the present. But there’s another way to approach the problem of time, and that is to ask how much the past is worth compared to the present. That’s the dilemma we face when questions arise about restitution for past injustices, such as past theft or slavery. In these cases, we must ask ourselves how much a past cost should count for in the present.

  One approach to restitution problems applies a standard rate of positive discount to the previous costs to convert them into a present value. Let’s say the discount rate is seven percent. If my ancestor stole a thousand dollars from your ancestor back in the year 1854, I should have to pay you seven percent compounded for each year between 1854 and 2018. Presumably that would undo this harm or rectify the injustice. But if you do the math, that amounts to more than $65 million. Most observers would correctly judge that such a restitutional award would be excessive. If the theft had not occurred, it is unlikely that you or your descendants would have had anything approaching that sum today. In fact, given how easily inheritances are squandered, you’d be lucky to have anything at all. Furthermore, no matter how loudly you proclaim the justice of this transfer, I don’t have the millions to give you. If we are going to award restitution at all, the amount of the initial theft plus some modest premium for suffering is more realistic.8

  The upshot is this: when we work from the past toward the present, we generally don’t—and shouldn’t—apply a standard rate of positive discount to make the later period of time less valuable. Again, a positive rate of discount cannot be applied uncritically, especially over long periods of time.

  Yet another argument pushes us toward relatively low discount rates. Not all values are economic values, and the pluralist bundle will likely contain what are called ideal goods, merit goods, and possibly Platonic values. By definition, these goods are not derived from individual preferences; rather, they refer to something which is objectively good for its own sake, according to some moral theory. For instance, when it comes to the value of Platonic beauty, we care that it exists at all; we care less about exactly which consumers have access to it when. Non-preference values seem to be the least eligible candidates for strong temporal discounting, and so a concern for non-preference values will strengthen the weight we should place on the future relative to the present.

  Maybe you’re not persuaded by these arguments. It’s difficult to ask serious economic and philosophic questions about the world five hundred years from now, or even a hundred years from now. The Einstein’s theory example was weirder still. Even talking about time preference across generations is confusing. So we don’t have a clear argument that the discount rate for human well-being should be exactly zero. But still, some significant holes have been punched in the idea that we should discount the distant future at very high rates and thereby assign it much less importance in our calculations.

  The opportunity cost argument

  Some economists offer the opportunity cost argument for discounting resource flows or dollars, and forgive me for a page or two because this becomes slightly technical.9 This argument notes that investing
a dollar today at positive returns will yield more than a dollar next year. Capital is generally productive. So a dollar today is worth more than a dollar in the future because the dollar today can be invested for more than a dollar tomorrow. In economic terms, the positive interest rate equalizes marginal rates of substitution over time, or, in other words, it expresses the value of a future dollar relative to a current dollar. This argument does not require positive time preference.

  This argument, by the way, does not suggest that we accept observed market interest rates uncritically as a measure of how much we should discount the future. We must adjust market interest rates for risk, transaction costs, and other complicating factors, such as taxes. Still, the market rate of interest would be a rough starting point for thinking about how much to discount the future.10

  The opportunity cost argument expresses a powerful logic, but, if understood properly, it does not militate against caring deeply about the distant future.

  First, the economic arguments deal with wealth only; these arguments do not establish strong positive discounting across well-being. If a higher rate of sustainable growth boosts well-being over time, it’s the well-being we care about. And we shouldn’t discount future well-being, even if we see reasons for discounting material resources or future income streams. Well-being is an end in itself, but wealth is a conduit to well-being and other plural values. For the most important part of the pluralist bundle—well-being—our deep concern for the distant future continues to hold. If we can produce ongoing gains in human well-being for the future, we have our Crusonia plant, and the case for maximizing sustainable growth remains.