A recent NBER working paper by a distinguished team of economists argues that a properly designed carbon tax can be a generational win-win. The team, led by Laurence Kotlikoff of Boston University, also includes Felix Kubler of the University of Zurich, Andrey Polbin of the Russian Presidential Academy of National Economy and Public Administration, Jeffrey D. Sachs of Columbia University, and Simon Scheidegger of the University of Lausanne.

By a generational win-win, Kotlikoff et al. mean a policy that would benefit not only future generations, who would reap the benefits of reduced warming, but also those of us who would begin paying the costs of mitigation now but would live to see only small, initial, climate improvements. The perception of a long lag between investments in climate mitigation and their full benefits has been a serious impediment to effective climate action. That is true both for democratic governments and for more authoritarian regimes, to the extent they are sensitive to public opinion. Although the paper discusses only carbon taxes, similar issues are raised by cap-and-trade, public investment, direct regulation, and other mitigation strategies.

The first section of this commentary outlines the Kotlikoff plan. The second section compares it to alternative strategies for dealing with the objection that climate action would pay off only in the distant future. The next section discusses whether the plan can properly be understood as an intergenerational “redistribution” in which the future “subsidizes” the present, as the authors contend. The concluding section examines the political realism of the plan.

The Kotlikoff plan

The difficulty of getting to an intergenerational win-win on climate policy arises from the fact that both climate change itself and mitigation measures operate over a long time-horizon. To simplify only a little, the adverse effects of global warming are an increasing function of the total amount of greenhouse gases (GHGs) emitted into the atmosphere  (see this piece from my colleague Joseph Majkut for a deeper explanation), while carbon taxes affect only the rate of emissions. Since GHGs, especially carbon dioxide, persist for long periods, even a big change in the rate of emissions affects climate damages only partially and over a long time period. Consequently, the ratio of the benefits of a carbon tax to its costs is less obvious in the short run than the long. 

That is especially true for the basic scenario examined by Kotlikoff et al., which is built on the work of Nobel Prize winner William Nordhaus. Under that scenario, if no carbon tax is imposed (“business as usual”), global temperatures would rise by about 4 C by the end of the century. Global GDP in 2100 would be about 4.7 times higher than it is now—but that would make it about 6 percent smaller than it would be with no climate damages. The acceptable degree of warming is higher and the resulting economic damages lower than those for the scenarios considered by other researchers—a point we will return to shortly.

Based on the assumptions of the baseline scenario, Kotlikoff et al. recommend a carbon tax that begins at $30 per ton of CO2 and increases at a rate of 1.5 percent per year. According to their model, if such a tax were imposed in 2020, only people born in 2057 or later would experience net benefits over their lifetimes. As Robert P. Murphy has pointed out, the decision to impose the tax in 2020 would have to be made by voters and their representatives born in the 20th century, but the benefits would accrue not to themselves or their children, or even, in most cases, to their grandchildren, but only to their great-grandchildren and even later generations. 

To make the carbon tax into a win-win, Kotlikoff et al. recommend rewarding early generations with a lump-sum cut in net taxes that would allow them to increase their consumption during the years after the carbon tax is first imposed. By a lump-sum tax cut, they mean one that is distributed equally to all citizens without regard to their level or source of income. In that regard, their tax cut would resemble the “citizens’ dividends” proposed, for example, by the Citizens’ Climate Lobby and the Climate Leadership Council. However, unlike those proposals, which are tax-neutral in the sense that total dividends equal total revenues from the carbon tax, the net lump-sum tax cut envisioned by the Kotlikoff model would require a “super dividend” in the form of a distribution or tax cut that would give more back to households than the revenues of the carbon tax itself.

The super dividend would be financed by issuing debt that would be retired by a tax increase scheduled for later years, after the reduction in climate damages was sufficient to produce net benefits. For example, Kotlikoff et al. calculate that in their baseline case, the net tax cut would give people born in 2020 a boost of 1.17 percent in their lifetime consumption, while those born in 2075 would be asked to pay extra taxes equal to 1.36 percent of their lifetime consumption. Even so, the scheme would produce a generational win-win since the burden of higher future taxes would be more than offset by the benefits of reduced climate damages. Importantly, the model assumes that utility is a function only of lifetime consumption. No allowance is made for utility received from direct enjoyment of environmental amenities.

In what follows, I will refer to these three elements — a carbon tax, a debt-financed net tax cut for early generations, and corresponding tax increases for future generations — as the Kotlikoff plan.

A comparison with other ways of getting to a generational win-win

Kotlikoff and his co-authors are not the first to grapple with the problem of persuading the current generation to adopt climate policies that have benefits that spread far into the future. This section compares the Kotlikoff plan to several alternatives.

A focus on stewardship. One approach is based on persuading people now living to focus on moral responsibilities toward future generations. The Christian climate stewardship movement is one example. As a climate scientist and a Christian, Katharine Hayhoe puts it this way: “We humans have been given responsibility for every living thing on this planet, which includes each other. We are called to tend the garden and be good stewards of the gifts that God has given us.” Those sentiments are shared no less strongly by secular climate activists like Greta Thunberg, who, in a speech to the United Nations Climate Action Summit, warned that “the eyes of all future generations are upon you. And if you choose to fail us, I say: We will never forgive you.”

Economists, on the other hand, are not traditionally comfortable with the language of moral responsibility. They are more inclined to search for ways in which self-interest, which they see as the most reliable predictor of human behavior, can be harnessed for the greater good. Most economists today, to be sure, are willing to expand the concept of self-interest beyond mere “selfishness” and to recognize that most people do care about the welfare of others around them. Still, they are inclined to reduce even altruism to a framework of individual utility. As an entry in the Handbook of Economics and Ethics puts it, 

[Individuals] follow their self-interest, even when satisfying other-directed preferences. In other words: altruism exists only when it increases the utility of the altruist to a larger extent than self-directed actions would have done, with given prices and constraints.

Working within that tradition, Kotlikoff et al. use an economic model based on overlapping, selfish generations. “Selfish,” as they use the term, does not connote any moral judgement, but simply means that in their model, individual welfare (or, to be technical, the model’s “utility function”) depends only on each person’s own lifetime consumption. They contrast this approach with other models that assume a benevolent and all-powerful social planner or a society of altruistic or infinitely-lived individuals. We will return to the issue of selfishness vs. altruism in the final section of this commentary.

Unused resources. A second proposed path to a generational win-win is to mobilize unused resources for climate mitigation. An example can be found in a Levy Economics Institute paper, “How to Pay For the Green New Deal,” by Yeva Nersisyan and L. Randall Wray. Writing from the perspective of modern monetary theory, the authors argue that “in the neoliberal era, we chronically operate below full employment.” So even with the official U.S. unemployment rate at a 50-year low, they see “substantial excess capacity.” They acknowledge that all-out decarbonization efforts might encounter bottlenecks that would require taxation or rationing to redirect key inputs from other uses, but they think unused resources would be able to do most of the job.

The Kotlikoff plan, in contrast, does not depend on unused resources. Instead, it is based on a model in which all available capital, labor, and natural resources are smoothly converted into useable output. The model is further simplified by viewing the economy in global terms, so that there are no imports or exports, and by consolidating personal consumption and government consumption. That leaves just two categories of final output, consumer goods and investment. The model allows total output to vary in response to changes in technology, induced, for example, by policies that favor clean energy over fossil fuel. For given resources and technology, however, consumption can increase only if investment decreases, and vice versa. 

In the Kotlikoff model, then, the increase in lifetime consumption that early generations get as a result of the tax cut does not come from a Keynesian boost to real GDP. Instead, it reflects a shift in the composition of GDP from investment to consumption.

A double-dividend. While on the subject of growth, it is also worth comparing the Kotlikoff strategy with policy prescriptions that aim for a “double dividend” by directing revenues from a carbon tax toward cuts in taxes that are believed to discourage saving and investment, for example, taxes on corporate profits, capital gains, or personal incomes. If such a strategy worked (and not everyone believes that it would), the payoff would include both lower carbon emissions and faster growth driven by greater investment. Conceptually, if the double-dividend effect were strong enough, growth could accelerate so much that early generations would experience increases in both income and consumption over their lifetimes, thus producing a generational win-win.

The Kotlikoff model differs from the double-dividend model in two fundamental ways. First, selective cuts to taxes on incomes or profits would violate the lump-sum principle that is baked into the Kotlikoff model. And second, viewed in static terms, the immediate effect of the double-dividend model would shift the composition of GDP toward greater investment, whereas the Kotlikoff model shifts the composition of GDP toward greater consumption.

Reexamine the costs and benefits. Still another way to get to win-win is to reexamine the scenarios that determine just how large a share of the benefits of their own climate actions the current generation might capture. For example, it has become popular in some circles to issue warnings that we have just twelve years to do something before climate catastrophe strikes. With a cascading rate of damages for global temperature increases beyond 1.5 C, not just the costs but also a good part of the benefits of an all-out effort to get to net-zero carbon emissions by 2030 would be captured by people now living. If we were to include co-benefits of reducing GHG emissions, such as reductions in local air pollution, which are disregarded in the Kotlikoff model, near-term gains would further increase. No intergenerational altruism or compensatory taxes and dividends would be needed.

As discussed above, the baseline scenario for the Kotlikoff plan posits a very protracted evolution of costs and benefits. However, the team does consider alternative scenarios that would increase near-term benefits. For example, in one scenario that is more pessimistic than the baseline, they estimate that the optimal carbon tax would increase to $70 and net benefits of a tax imposed in 2020 would be felt by people born as early as 2024.

Resources from the future. One final idea that perennially pops up calls for using resources borrowed from the future to finance decarbonization today. But unfortunately, although we can preserve and enhance resources for the future, as the aptly named research organization urges us to do, we can’t draw on physical resources from the future for current use. 

Ludwig von Mises made this point forcefully in a 1918 lecture on financing Austria’s participation in World War I: 

Some people claim that financing the war by state loans is tantamount to passing on the costs of war from the present generation to future generations. It is sometimes said that this transfer is fair because war is waged not only in the interest of the present generation, but also for that of our children and grandchildren. Nothing could be further from the truth. War can be waged only out of currently available goods. One can fight only with the weapons on hand; all military needs must be met out of existing wealth. It is the present generation that is waging war from an economic point of view, and it is this generation that must bear all the material costs of the war. Future generations are affected only insofar as they are our heirs. We will be leaving less behind for them than if war had not happened. This is an unavoidable fact, whether the state finances the war by indebtedness or by any other means.

Just as the Austrian soldiers of 1918 could not fight their war with artillery shells made in 1928, we cannot fight our war on climate change with wind turbines manufactured in 2057. We can, as von Mises pointed out, affect the welfare of our heirs by our decisions to use our present resources destructively, say by fighting wars, or benignly, by investing and conserving, but they can do nothing for us. 

The Kotlikoff plan does not rely on any movement of real resources from the future to the present. However, some of the language used in the paper is, I think, misleading in that regard. The next section discusses how the relationship between the present and future in the Kotlikoff plan should properly be understood.  

Present and future in the Kotlikoff plan

As we have seen, the Kotlikoff plan uses debt-financed tax cuts for early generations balanced by debt-retiring tax increases for later generations to ensure equal utility gains for all generations. Implementation of the plan is assigned to a hypothetical administrative body called the Lump-Sum Redistribution Authority (LSRA). The work of the LSRA is described in these terms:

Our LSRA mechanism can be viewed as a nondistortionary deficit policy in which the government cuts taxes in a lump sum manner for early current and newborn generations based on the figures in the table and then raises the net tax, again, as indicated, to service the associated debt. (p. 27)

Kotlikoff invites us to think about the plan as a Pareto-enhancing exchange: 

Carbon taxation, coupled with appropriate intergenerational redistribution, can make all current and future generations better off. Indeed, it can make them uniformly better off. (p. 3) 

Calculating Pareto improvements is standard procedure for determining optimal policy responses to negative externalities.  … Ours appears to be the first large-scale study of Pareto-improving carbon taxation. (p. 4)

 Achieving these win-win outcomes does, however, require significant redistribution from future winners and transfers to early losers. (p. 41)

Sharing efficiency gains evenly requires, however, taxing future generations by as much as 8.1 percent and subsidizing early generations by as much as 1.2 percent of lifetime consumption. (Abstract)

That language makes it sound like the LSRA is facilitating something like the exchange Ronald Coase had in mind when he discussed bargaining between a railroad and farmers whose crops are damaged by sparks from passing trains.  Coase identified one possible outcome of those negotiations as an agreement by farmers to pay the railroad to install spark suppressors on its engines. Just substitute our descendants for the farmers, ourselves for the railroad, and a carbon tax for the spark suppressors. 

With a little thought, though, we can see that the resemblances between the Kotlikoff plan and the farmer-railroad bargain are only superficial. For one thing, the Kotlikoff plan involves no actual agreement. We can’t communicate with our descendants to learn their preferences or secure their assent. We think they will be grateful to us if we start diverting capital and labor now to mitigate global warming, but anything we do must be based on our judgment, not theirs. 

More importantly, even if we could communicate with our descendants and bind them to an agreement, there is nothing—literally nothing—that the future can do for us. Unlike the farmers in the Coasean example, the future can neither give us, who stand in the place of the railroad, the labor and capital we need to build our spark suppressors, nor can they compensate us in any other way if we use our own resources to build them. 

Because of these crucial differences, it is important to be careful when using words like “redistribution” or “subsidy.” In a contemporaneous transaction like the farmer-railroad example, “redistribution” refers to a process in which one group, the farmers, transfers physical or financial resources to another group, the railroad, in return for a reduction in externalities. Such a transaction could plausibly be described as a “subsidy” of the railroad by the farmers in the sense that the resources relinquished by the farmers are the very resources acquired by the railroad.

In the Kotlikoff plan, the words “redistribution” and “subsidy” have completely different meanings. Yes, carbon mitigation undertaken by early generations does produce real environmental benefits for later generations, and yes, the plan does envision an increase in the consumption of early generations and a reduction in that of later generations. However, the taxes paid by the later generations to retire previously issued debt can in no way be construed as the source of any benefits received by the earlier ones. Also, lower consumption by later generations is not a result of the taxes they pay to retire the bonds issued in the early years of the plan. Those payments are merely financial transfers within their own generation. Instead, future generations have less to consume because we left them a stock of capital that was more heavily invested in production of clean energy and less heavily invested in production of consumer goods.

In fact, nothing that happens in the future can be said to cause anything that happens in the present. Among other things, no harm would come to anyone in the present were future generations to decide not to carry out their end of the bargain, say, by continuing to roll over the bonds we issue now instead of retiring them, or even simply by defaulting. Taking all this into account, it seems fanciful to refer to taxes that may (or may not) be paid in the future as the source of a “subsidy” to present-day consumers. 

Similarly, the increase in current consumption envisaged by the Kotlikoff plan would not be the result of any “redistribution” in the sense that, say, a tax on New Yorkers to finance Pell grants for students in Alabama redistributes wealth between those states. Instead, any increase in the use of resources for current consumption can only occur if there is a reduction in the use of resources for some other current purpose—as if, say, the state of Alabama cut housing assistance in order to finance the increase in Pell grants.

But, if the Kotlikoff plan is not a “redistribution” or “subsidy” between generations, what is it? A better way to understand the Kotlikoff plan, in my view, is to drop the authors’ premise that “generations are selfish” in favor of the more realistic assumption that each generation is composed of a mixture of “selfish” and “altruistic” individuals. I put “altruistic” in quotation marks here to indicate use of the term in the sense discussed previously, namely, to denote the preferences of people who care not only what happens to themselves, but also what happens to other people. The other people, in this case, are the yet-unborn inhabitants of the future.

We could, instead, speak of people who are “present-oriented” and “future-oriented.” To get technical, we could speak in terms of people with higher and lower discount rates. But I think it will be clear enough to use “selfish” and “altruistic,” without the quotation marks, as shorthand for the whole constellation of attitudes.

If our own generation did not have a mixture of selfish and altruistic members, the whole issue addressed by Kotlikoff et al. would disappear. If we were all selfish, and if a carbon tax would not produce net benefits in our own lifetimes, we would just leave future generations to deal with their own problems. On the other hand, if we were all altruistic, we would have enacted effective mitigation policies long since. 

The fact that some of us are more altruistic toward future generations than others is the whole driving force behind the debate over climate policy. The altruists want an effective climate policy, but they can’t get a majority on board. Properly understood, then, what the Kotlikoff plan proposes is a Pareto-improving bargain within the present generation, with altruistic members offering a payoff to selfish members in return for their support for climate action. The payoff is not delivered through a wormhole from the future, but rather, through a tax cut that diverts present-day resources from saving to consumption. The desired climate effects are then achieved by focusing a greater share of the smaller amount of saving on investments in decarbonization.

The required diversion from investment to consumption is exactly what we would expect from the imposition of a carbon tax with a rebate of the proceeds equally to everyone. The burden of the carbon tax would fall disproportionately on people with relatively high incomes, since they are responsible for a disproportionate share of emissions. Although the poor are responsible for more emissions per dollar of income, the rich, with their higher incomes, are responsible for more total emissions. Since the rich save more than the poor, the intragenerational redistribution of income would reduce total saving. Considering the “super-dividend” feature of the Kotlikoff plan, the effect on saving would be even more pronounced.

Some readers might object that the net tax cut, starting from a state of full employment, would cause inflation. Other things being equal, it might, but not if the decrease in saving were offset by an equal decrease in total investment. Making sure that would happen might require measures such as a revenue-neutral increase in marginal tax rates on investment or cuts in government net investment, exempting investment directed to climate mitigation. 

It might also be objected that a decrease in investment would hurt future generations, who are just the people we want to help. Other things being equal, it might, but not if it takes the form of an increase in climate mitigation while making an even larger reduction in emission-intensive projects. To a considerable extent, the carbon tax itself would accomplish the necessary realignment of investment, since low-carbon projects would become more profitable and high-carbon projects less so. Policymakers who are not market purists might want to add regulatory measures or direct government investment in green projects to reinforce the tax-induced shift toward clean energy. 

Clearly, smaller total investment combined with a shift toward mitigation efforts will make the future better off only if the rate of return on green investment is higher than the return on investment of other kinds. But if it is not, then there is no economic case for climate mitigation policy in the first place, and our whole discussion is moot. If it is, then a smaller total amount of investment, better-directed toward high-return, low-emission capital, will make future generations better off. 


This commentary has discussed whether a carbon tax can be made a generational win-win. The issue is important, since a tax whose costs are felt more quickly than its benefits is a hard sell.

The Kotlikoff plan aims to achieve a win-win using a carbon tax, a debt-financed net tax cut for early generations, and higher taxes on future generations to retire the debt. I have argued that such a plan is best understood as a Pareto-improving bargain in which “altruistic” members of the current generation offer higher immediate consumption to “selfish” members in return for their support for a carbon tax. Although future generations do not participate in the original bargain, they gain to the extent they inherit a capital stock that is smaller, but more heavily skewed toward clean energy, in line with their presumed interest in mitigating climate damages.

I find the economics of the Kotlikoff plan to be plausible, even in their baseline case, which posits moderate climate damages and a rather high maximum tolerable global temperature. However, I am less certain about its politics. It might turn out to be neither necessary nor sufficient to break the political deadlock over climate action.

The Kotlikoff plan might be unnecessary if, as discussed earlier, a majority of the politically active population became convinced that climate benefits and co-benefits within the lifetime of those now living would justify a carbon tax. That could happen if better estimates of the pace of climate change indicated a more rapid pace of warming than is now expected, or simply if the public became more concerned over what we already know. Growing public demand for action might also be driven, at least in part, by concerns over the impending loss of intangible environmental amenities, a consideration omitted from the Kotlikoff model. 

On the other hand, the Kotlikoff plan might turn out to be politically insufficient even if economic and scientific considerations favored it. The plan calls for an altruistic faction to persuade a selfish faction to consent to a carbon tax in exchange for an increase in consumption. That could happen only if the balance between the factions is such that neither can act by itself. If, instead, the selfish faction is politically dominant, why would it go along with the Kotlikoff scheme? Why not just grab the dessert of a massive tax cut without the spinach of a big increase in energy prices? In fact, that is just what seems to be happening recently, with economic growth driven by a rising share of consumption in GDP while decarbonization efforts languish.

The Kotlikoff plan might also turn out to be insufficiently attractive to climate activists. Backers of the Green New Deal and similar programs often speak in terms of putting the economy on a war footing to fight an impending climate emergency. They call for at least a slowdown in the growth of consumption spending, if not an actual decrease, and a corresponding increase in total investment. Such activists are likely to view the Kotlikoff plan as weak tea. The  Washington state carbon tax referendums, which were opposed by many climate activists as too timid, could be a case in point.

All things considered, then, the Kotlikoff plan may turn out to be little more than a theoretical curiosity. Still, its authors have made a useful contribution. Their work draws attention to the inconvenient truth that people who are rational, well-informed, and self-interested—the very people who inhabit the world as seen by economists—may hesitate to get on board the climate action express. Overcoming their resistance will require not just elegant models, but a lot of hard political work to back them up.

Photo by Markus Spiske from Pexels