This Is How It Gets Better

It’s a real privilege to comment on this book. From the Mars trilogy to my personal recent favorite, Aurora, Kim Stanley Robinson has been one of my favorite science fiction authors, staying with me as I went from teenage escapism to middle-aged escapism. There are so many great ideas in The Ministry of the Future (TMFTF), where Stan has clearly combed the academic and activist literature for the boldest ideas to grapple with the climate crisis and used the medium of fiction to communicate them. There are engineering feats, like the propping up of glaciers to slow melting, direct air capture of CO2 at economically feasible scale, alongside political transformations like the mutually-assured-destruction made possible by targeted kinetic pebble smartbombs, a rebirth of Indian democracy, and carbon quantitative easing? Everyone who cares about climate change (and really at this point it should be everyone with some stake beyond the next 10 years) should read it. But beyond being a hardware store full of tools for decarbonization, it also charts a politically possible trajectory to a transformed economy. TMFTF is not just outlining a future sustainable economy, but showing a properly historically contingent path to it.

Stan’s book is simultaneously an Edward Bellamy and an Edouard Bernstein for our times: a possible future economic configuration for humanity but also a way to get there that builds on, rather than replaces, the institutions and societies we currently have. The engine of the novel, of course, is the threat of civilizational collapse posing a stark fork in the road of human progress: eco-social democracy or barbarism.

For economists, the book is interesting both as a criticism of the view of climate change embedded in integrated assessment models as well as an expression of hope in the informal empire of central banking authorities. “Carbon quantitative easing” is the primary engine by which our characters move decarbonization at a world scale, harnessing markets to deliver rapid global adjustments and innovations in sequestration. The premise is ingenious: rather than leading with the (necessary) negatives of a carbon taxes, which have proven to be politically unpopular even when proposed at hopelessly inadequate scales (no matter how many petitions are signed by economists), policy leads with debt-financed de-carbonization subsidies, which create a whole new global economy organized around sequestration. Once it pays to keep and pull carbon out of the sky, the global economy pivots, with the kind of speed and coordination that we absolutely need, ones that would be hard to generate using purely dirigiste tools.

My particular take is going to be to use TMFTF to indict how economists have modeled climate change in the macroeconomy, both theoretically and empirically. But I’m also going to criticize the principal, somewhat reluctant, sidekick for saving the world: the central banks. Still, I think Stan is really onto something: this is how we get to something better and sustainable, not waking up the morning after a mass revolution and writing a new constitution while implementing some algorithmic central planning AI (or for libertarians a laissez-faire utopia), but rather piecemealing together a patchwork of mechanisms, some from above and some from below, and an equally Frankenstein-like quilt of political coalitions that together make the burden of scarce resources both less binding and borne more equitably.

The Economy Is For People

The opening scene of TMFTF is unforgettable, a close up description of a murderous heat wave, with the predictable armed conflict over water and electricity, hitting India. This heat wave kills 20 million people, precipitates an honest reckoning with climate change by both the formal powers, as the Indian government unilaterally throws reflective material into the sky to geoengineer the planet in the interest of its population, as well as informal powers (the Children of Kali terrorist group). The terrorists kill innocent civilians, albeit somewhat discriminately, injecting cows with cultured mad cow disease, blowing up coal plants, and knocking fossil fuel powered jets out of the sky with pebble bombs, and indeed, anti-carbon direct action does start yielding some progress on climate goods.

Temperature is already a silent killer in rural India. 40% of India’s 1.3 billion people live and work in the agricultural sector, and agricultural productivity (for obvious reasons like it is outdoor work and less obvious reasons of crop failure and water shortages) falls with extreme temperatures. Extreme temperatures do not kill yet in massive heat waves with millions dying in a few months, but slowly, as frogs boiling, with raw rural poverty interacting with extreme temperatures to kill infants and the old, thousands of tragedies that, over time and space, already constitute a scorching grim reaper. Burgess et al.(2017), find that rural Indian life expectancy strongly responds to temperature increases, and the magnitudes imply 5.1 years and 10.4 years of life expectancy lost for those born in 2045-2059 and 2075-2099, respectively. These are mostly economic effects, not accounting for the direct effects of temperature on human survival (because they are based on historical variation in weather which have not so frequently been in the lethal range).

Delhi, with a population over 11 million, experienced 3 weeks of 45 degree plus Celsius weather in 2019 that killed 36 people, and this was with the electricity on and the government functioning. In 2015, over 2000 people died during another heatwave in South India. Even small increases in mean temperature imply tail events that are increasingly fatal. Stan does us all a favor by pointing out that our fragile meat-and-water bodies are directly destroyed by high enough temperatures (coupled with humidity that prevents us from sweating), and drawing attention to the salience of tail weather events driven by secular climate change, where wet-bulb temperatures greater than 30 (higher than 50 Celsius) are lethal. In a decade or so, the Indian heat waves will likely hit this level, and the chaos and death vividly etched in TMFTF’s opening does not seem implausible.

It’s useful to contrast the opening of TMFTF with the bloodless specification of the “damage function” in integrated assessment models (IAMs), which lie behind estimates of the social cost of carbon. The sociology of these IAMs deserves greater study (and for all I know this has already been done). These models have four ingredients: i)a trajectory of carbon emissions based on projections of carbon intensity of production plus future GDP growth, ii) a scientific model mapping carbon emissions into temperatures, iii) a “damage function” mapping temperature into estimates of lower GDP, and iv) some social welfare function that applies a discount rate and a utility function to the resulting path of output. The derivative of the discounted social welfare function with respect to carbon emissions yields the marginal social cost of carbon, which is a number that gets used throughout government cost-benefit analysis (see here for how these numbers are estimated and used).

These models are a wonderful window into so many pathologies of economic policy modeling, from the insistence that the neoclassical growth model is the appropriate benchmark macroeconomic model (which assumes full employment, no role for fiscal or monetary policy), to the preservation of the global distribution of income to taking the U.S. economic parameters as global ones to the assumption that future generations out to be discounted based on the market rate of return.

The premise of a lot of economic modeling of climate change is that it involves an intertemporal choice between emissions today and climate change borne by future generations. There are many philosophical objections to discounting future generations, and much debating about what discount rate to use (e.g. if the market discount rate is set by people who have finite lifespans, is it really the right one to use? Should any discount rate be used?). In my opinion, we are already past this decision point: the costs of climate change are here and being borne by an increasing share of the world’s actually already existing population, which is overwhelmingly in poor countries. And the benefits of emissions accrue roughly proportionally to income, so the airplanes, food, and infrastructure of rich people and rich countries can account for the lion’s share of both the past stock and current flows of emissions (see evidence by Chancel and Piketty here). Yes China and India are major emitters, but it is also important to flag the enormous inequality in income, and emissions, even within China and India.

Consider an early assessment by Nordhaus (1993). The DICE model assumes that a 3C warming would lower world output by 1.3 percent. In light of the past decade of heat waves, forest fires, hurricanes, floods, and generally increasingly treacherous natural environment, the ever-increasing evidence of anthropogenic climate change, and the mortality estimates above, this seems maddeningly optimistic, and a future generation may wonders whether the Hague would have been a better destination for the author than Stockholm. I rarely get so upset as when I’m discussing the role economists played in the 1990s and 2000s (and some even to this day) in minimizing the threat of climate change and suggesting that we could and should all live with 3.5 degree increase, as GDP growth and innovation would allow future generations to adjust easily to any costs.

I was going to write a lot more here, but I see Noah Smith beat me to it, in a post I completely endorse. The last decade has seen an explosion of climate economics that finally gets a sense of the threat (to be fair only about a decade behind the natural scientists), but the legacy of economists waving off the warnings of scientists is going to be hard to live down.

The most consequential effects of climate change on human well-being is probably not summarized by a GDP penalty, it is direct biological effects on human bodies, magnified by direct physical impacts on physical infrastructure and food supply chains (which are small shares of GDP but obviously crucial for welfare). Imagine we get heat waves (and hurricanes and forest fires and maybe even epidemics) that raise mortality a little tiny bit: that human mortality increases by 0.01%, or one over ten thousand, as a result of climate change. That’s 700000 people killed directly by climate change a year (modestly greater than U.S. deaths from COVID so far). Now multiply that by roughly 5 million dollar value of a statistical life (ignore the problems with that number!), and you get an estimate of damages that’s around 3.5 trillion, already 4% of world gdp, without accounting for any of actual economic effects created by destroyed infrastructure and agriculture. Add even a little bit more mortality, and you get costs of climate change that swamp the world economy.

One doesn’t even need to go down this particular rabbit hole to get quite high social costs of carbon. Stern and Weitzman argued for damage functions that were steeper than quadratic. Weitzman convincingly argued that the distribution of damages was fat-tailed, and if you push this logic far enough you can get that the costs of climate change are basically infinite. What use is the language of trade-offs and social optima when faced with singularities in the path of possible economic trajectories?

The point of this is not that there is a better damages function, but that there is something deeply wrong in the belief that the environment’s effects on human well-being could be summarized by its effects on the economy. Or that there is some entity called “the economy” that somehow exists apart from the humans that live in it. The hubris of economists was that we could trade-off the environment with economic well-being; but such a trade-off implies a concave objective function, where well-being today can be smoothly traded-off for well-being tomorrow. If instead we have these non-linear feedbacks and freakish ecological and political effects of climate change, the language of trade-offs and cost-benefit analysis seems of limited use: we’re in do-or-die territory and the preconditions for a social trade-off are gone.

TMFTF inverts the principle of the IAM. Instead of seeing the environment as a input into the economy, it sees the economy as something that needs to be subordinated to the “throw everything at decarbonization” mandate that follows from managing a runaway nonlinear process that threatens not just the vast mass of biological life, but also the accumulated physical infrastructure and slowly depreciating investment goods of the last few centuries.

The Empire of Central Banking

The trump card in TMFTF is the carbon coin, where the central banks of the world agree to issue a currency to compensate agents who remove carbon from the atmosphere. This can include leaving proven reserves of fossil fuels in the ground, or figuring out new ways of pulling carbon out of the atmosphere. This is a great idea, although the real work is obviously figuring out how to cheaply and accurately monitor pinpoint sequestration efforts. Once we have the enforcement technology and administrative capacity to monitor detailed emissions and emissions reductions (as well as property rights around who should be paid!), we can create a variety of devices to provide decentralized incentives to both abate and sequester carbon. TMFTF goes with the global monetary policy and financial system bank shot, instead of, for example, a much more workaday system of cash payments.

One view is to forget the baroque mechanisms of currency exchange markets setting the relative price of carbon coins and local currency: imagine instead that governments just agree to pay for sequestration. This is simple, and could have been put into the American Jobs Plan: prove to a regulator that you’ve sequestered X tons of carbon and you get Y dollars in actual USD. Economists enthralled with the carbon tax could be equally enthralled with the sequestration subsidy (of course conditional on the monitoring problem being solved). Anything that alters the relative price of carbon vs non-carbon sectors is basically equivalent when you see everything as a supply-and-demand diagram, and future generations should gladly accept whatever debt burden (if any) it takes to pay off current emitters, so why not do it with a decarbonization subsidy rather than a carbon tax. The politics of implementing a sequestration and anti-emission subsidy seem far less like science fiction than an adequate carbon tax, at least right now. Buying off the key fossil fuel actors (fossil fuel companies and Gulf states, although I would loved to have seen speculation about Venezuela’s position) and creating a financial sector interest in managing climate change seem to be the major political obstacle, which is overcome by subsidizing de-carbonization rather than taxing carbon. These are important, but it’s not clear to me there are any actual administrative or economic merits to running decarbonization through the financial sector. Every doomed cap-and-trade policy had to provide some carve-out for incumbent emitters, and indeed the redistributive conflicts over those carve-outs arguably sunk some proposals (e.g Waxman-Markey). TMFTF accepts that on the time-scale we face, we need to buy-off the carbon interest rather than politically conquer it. But this has always been a tension in the eco-socialist program: if climate change is upon us, and is a species-threatening event, shouldn’t we put the class war on hold and encourage capitalists to chip in with their eldritch command of innovation, arbitrage, and technology to save us from it? To which a response is: fossil fuel interests can’t be bought off, and thus must be politically defeated by a large climate coalition brought together by material and environmental interests, hence the political necessity of a Green New Deal platform. I’d be curious as to Stan’s take.

One might wonder (and I did at first) what the advantages of a new financial asset are in delivering this global subsidy that incentivizes sequestration. I think it is primarily political and administrative: a global consortium of the major central banks is the only global economic governance tool we have right now, and climate change requires something both effective and global. Almost every subnational carbon tax has run into political problems, and one possible reason is that everyone gets the free-rider problem inherent in such piecemeal solutions: why do we have to suffer these taxes while they continue to emit? And given we’re not getting a global tax authority anytime soon (although Janet Yellen and the Treasury might be working on it), TMFTF looks to the one global economic power we have that have the scale and reach necessary to implement global decarbonization: the network of coordinated monetary authorities of the large economies, documented for example in Adam Tooze’s Crashed. The consortium of central banks swayed by Mary’s repeated injunctions that climate change falls under their mandates is an extremely non-democratic entity staffed by eminently reasonable people; the closest humanity has ever gotten to the economist ideal of a social planner.

It is true that coordinated large economy central banks are possibly the only apparatus we have with the ability to shape the whole global economy and it is at the same time revealing of the limits of our political imagination (not Stan’s)! One of the legacies of the 2008 financial crisis and the innovations in monetary policy pursued around the world was the increased interest in money and finance, with a lasting consequence being that the monetary system became an intellectually interesting institutional matrix by which we could backdoor implement a variety of policy ideas. It led, in some corners of the left, to a kind of Green Lanternism of the central bank, where Federal Reserve balance sheet innovation is the primary tool we need to fix any economic problem. We were going to fix wealth inequality with Fed deposit accounts, fix unemployment by having the Fed stand ready to finance any deficit spending, and in TMFTF, fix climate change by issuing a new financial instrument.

None of these are obviously wrong ideas, but it’s worth wondering why people want to lean on the financial system and central banking to do all the work of economic governance. I suspect some of it is because mastering the details of the financial system is a certain kind of esoteric expert knowledge, like subgenres of music, mastery of which generates joy among some people. Some of it is also that money has a semi-legal, semi-conventional character, where it is both created by tax liabilities of the state but also private notes of exchange and financial contracts, and because of the numerous tangencies between taxes, law, regulatory authorities and the administration of money the crucial central bank role in financial markets becomes a powerful lever for governments to use in redistributing income and reshaping the economy. Finally, it is a relatively politically autonomous domain, deliberately shielded from party politics. Experts and interests can then offer ideas for interventions within a powerful, politically insulated institution to secure their desired outcomes, be they pecuniary or ideological. So climate activists just have to convince the Fed governors, not voters.

TMFTF is quite open about the non-democratic, somewhat magical nature of central banking. Part of the power of central banks, in TMFTF and in real life, come from their perceived ability to make promises that are insulated from politics, guiding investment decisions of private actors, and institutionalizing this promise-making capacity was essential for central bankers to be able to issue pronouncements that could coordinate the private markets. Central banks that move too far and too fast from the expectations of private financial actors can find their ability to influence the economy undone by either price-setting or currency flight or both. On the flip side, central banks that maintain the power of the spell gain capabilities to guide the economy via forward guidance, where the economy can be steered merely by credibly signaling future policies. At the extreme, you get the argument that a central bank that pledges to wreck the economy should inflation get too high will never face substantial inflation. For some economists, technocrat-plus-banker capture of central banks is the political precondition for central bank efficacy. The influence of the central bank on private financial markets is contingent on it using it in the interest of the system as a whole, regardless of what elected legislatures demand. But given the crises unleashed by climate change documented in the book, why would states continue to give the banking authorities such long leashes? Why are the bankers able to do this kind of autonomous response to crisis? What stops the central banks from changing their minds later on and reneging on the carbon coin?

One answer is raw political power, either from above or from below. One could imagine a global green version of Fed Up showing up at the houses of the bankers with placards and protest songs, and this mobilization being the pressure that keeps the central bankers on their mission. One can also imagine legislatures transforming the governance of the central bank so that “Federal Reserve Governors For the Future” institutionalize a climate interest in central bank governance.

The central banks are just the cavaliers of the winning army in TMFTF. The supporting protagonists of TMFTF, from Mary’s team at the Ministry to Jane Yablonski at the Federal Reserve, are all technically competent bureaucrats, democratically unaccountable, loathed by many, bureaucrats, who save the day. Bureaucrats who have discretion, who do not have their every move watched and policed by Congressional committees and oversight reporting (Mary, iirc, never has to fill out paperwork or write a report to the UN!), and who are far outside the iron cage of administrative roles.

Making saviors out of bureaucrats is a worthy rhetorical goal, for we will need public-spirited bureaucrats to save us. I look at all my colleagues and friends who have entered the Biden administration and I’m somewhat envious of their confidence that they can really make a difference in government. But many of them also recognize the historical moment created by COVID-19 and the mobilizations of Black Lives Matter in the past year have created an activist base that can put outside pressure on the clunky apparatus of Federal government that strongly complements insider pressures. But both bureaucratic politicking and street protests exercise power through minoritarian democratic channels, not majoritarian ones. Surely a democratic system would have ways of aggregating a collective interest in a more direct fashion. When Mary is visited in the night by the traumatized survivor of the Indian climate massacre, it’s an admission that democratic, mutual persuasion based means of securing a climate transition is impossible. Strange men living in sheds handing out midnight threats is no basis for a system of government!

TMFTF is a realistic projection of the cramped and depressing politics our time, and so the novel isn’t quite utopian. But it is funny that the only systematic force that keeps bureaucrats accountable and responsive to public pressures and climate needs are their own consciences, supplemented by terrorist attacks and midnight visits by traumatized radicals. It seems easier to imagine the end of ecologically-destructive capitalism than it is to imagine actual democratic governance at the scale needed to address the climate crisis.

The Path-Dependent Transition

The final point I’d like to make is that the historical trajectory Stan charts for our future is a winding one. Social and technological innovations happen that were unforeseen, political configurations and interests evolve, and all of these shape and constrain the institutions that emerge. The 10% rational policy, 90% jury-rigged nature of social problem solving (particularly at the global scale) comes alive in the novel. It still winds up working in the end, but it is as much by chance as anything else.

This is usefully constrained with the “Red Monday Blues” fantasy that perhaps still lurks in the lizard brains of many leftists. Basically the working model is that there is a revolution or decisive election, and then we need to get to the business of implementing a socialist economy. And so ideas for whole cloth systems of planning, from parecon to some kind of nationalized Amazon-mondragon hybrid, to sui generis models of market socialism with various Rube Goldberg devices thrown over markets to discipline them, proliferate. These visions all suffer from the Rawls “ideal theory” problem: they presume some break, some moment that will come that will let us re-engineer political and economic institutions from the ground up.

But TMFTF instead charts a much more historically plausible path forward. Where idiosyncratic legacies of the past exercise enormous and difficult-to-predict effects on the future, so we never get clean breaks, elites are never permanently dislodged, and revolutions are always only partial breaks with the old order. Past actions and structures deal the cards held by the living.

Social democracy, a civilizational achievement by any light, was clearly pulled together like this. It was considered a second-best compromise by electoral socialists who nonetheless pushed for it, but was also made possible by simultaneous innovations in administrative capacity (e.g. automated payroll systems and ever-larger firms) and a couple of global conflicts bracketing big demands for increased social insurance. Nobody would look at the architecture of the welfare states and conclude that they were elegant expressions of abstract, parsimonious principles of social scientific genius. Instead they are muddled-through mixtures of utopian ideas, inspired technocratic hacks, bureaucratic innovations, and political deals made between factions of politicians and fractions of the population. Similarly with neoliberalism, which for all its efforts never really succeeded in reversing Wagner’s law (that government is a normal good that increases with share of GDP), and wound up disappointing its most fervent adherents.

Stan’s road out of climate crisis has a similar view of path dependent transitions. Instead of replacing the whole economy with something completely different, it builds on the enormous infrastructural and institutional power that modern civilization, including those realized in varieties of capitalist arrangements, harnessing them to deliver something on the other side of terrible. But the anticarbon economy winds up unleashing a whole slew of new economic mechanisms, gestured at throughout the book, like parecon and mondragon. The vision embraced by Stan’s book looks much more like a yin and a yang of central direction and bottom-up initiatives. The central banks provide a sparse guiding light for human economic activity, but the vast bulk of economic activity is still something autonomous and bottom-up, if not purely market driven, that actually executes the decarbonization at the required scale.

What I think TMFTF gets right is the bricolage nature of the economic calculation problem. It is neither a pure Hayekian vision of prices bring privately known tastes and technologies into harmony nor a Lange-ean vision of government calculation of social optima, to be implemented by an optimal tax structure or government price setting, but rather some fractal combination of the two, with levels of centralization nesting decentralized organizations that in turn nest even smaller centralized organizations. Coase drew a binary line between command inside the firm and prices outside of it, but maybe its more useful to think of the economy as nesting multiple scales of bottom-up signaling and communication and top down means of coordination and aggregation.

Indeed, I wonder if modern states already have the tools they need to radically reshape the economy, with limits set by norms and politics. Beyond central banking setting bounds on the price of capital, we have large chunks of the labor market regulated by minimum wages or directly employed by the state, we have public provision of schooling (rationed using mechanisms Soviet planners didn’t even dream of), and a detailed and sophisticated tax and transfer system. The idea that we need some radically new tools of government price setting in order to really get government control of the economy is basically belied by a quick inventory of the tools of government regulation we already have. But because we already have them, it reveals that the obstacles to democratic planning are not economic or administrative ones to incentive-compatible, information-constrained planning, but an absence of models of democracy that we trust absolutely enough that we could rely on them to run the economy. The planning problem is not an economic design problem: it is a political one. One that can’t be solved in some transcendental way but that requires us to work with the wildly varying ideas and interests and values of the people and groups we have right now.

Suppose we massively change the relative price of carbon while driving down the costs of clean alternatives (which we may well do) in the next decade or two. We will risk many political and economic effects that are difficult to anticipate, and the responses to these effects will likely be improvised and the product of contingent coalitions, interests, and ideologies. Paying an enormous sum for sequestration efforts will transform (as is the point) huge sections of the economy. We should probably admit that our social science isn’t good enough to predict the detailed contours of the outcomes if we do this, even as our climate science is good enough to predict the broad calamity if we don’t.