7 WHY IS GREEN INVESTMENT SO DIFFICULT?

Bob Hancké, PEACS

25 February 2025

The green transition will not be cheap, at least not in the short run. In a report from early 2022, the McKinsey Gobal Institute estimated for it to run to $275 trillion for the 30 years between 2021 and 2050, the provisional date by which the world is scheduled to reach net zero:

‘Capital spending on physical assets for energy and land-use systems in the net-zero transition between 2021 and 2050 would amount to about $275 trillion, or $9.2 trillion per year on average, an annual increase of as much as $3.5 trillion from today. To put this increase in comparative terms, the $3.5 trillion is approximately equivalent, in 2020, to half of global bcorporate profits, one-quarter of total tax revenue, and 7 percent of household spending. An additional $1 trillion of today’s annual spend would, moreover, need to be reallocated from high-emissions to low-emissions assets. Accounting for expected increases in spending, as incomes and populations grow, as well as for currently legislated transition policies, the required increase in spending would be lower, but still about $1 trillion. The spending would be front-loaded, rising from 6.8 percent of GDP today to as much as 8.8 percent of GDP between 2026 and 2030 before falling. While these spending requirements are large and financing has yet to be established, many investments have positive return profiles (even independent of their role in avoiding rising physical risks) and should not be seen as merely costs. Technological innovation could reduce capital costs for net-zero technologies faster than expected.’ (McKinsey GI, The net-zero transition: What it would cost, what it would bring, January 2022)

The reason for this daunting up-front outlay is simple: we need to replace practically every aspect of our fossil fuel-based economy and we will end up doing almost the same things as we do today, but with different means. There is no (or very little) short-term gain, in other words. I will discuss this question of the macroeconomic cost in more detail a few weeks from now. Even there are, I think, several mitigating aspects that make the bill more palatable than many pessimistic observers suggest, the costs will be significant regardless. In this post, I want to concentrate on green investment, and the role that uncertainty and accounting conventions play in that.

Let's start with the raw figures. While the cost of the green transition is high, these costs are not necessarily a problem, as many large infrastructural projects around the world, from the Channel Tunnel over new towns in the Arab Sea to the London and Paris Olympics suggest. Moreover, the world cannot really be said to be short on cash: pension funds, asset management funds and other large-scale saving and investment schemes currently hold multiples of annual world GDP in assets. Yet, despite all that, funding will likely be a problem in the green transition. The key reason is that while investors can quantify risk, they cannot directly address the uncertainty that defines the green transition (and which I discussed in my previous post). Even though there is no doubt that some individual schemes may have a very specific narrow risk/return profile that attracts funding, in aggregate, green projects, which require a lot of necessary but ‘non-performing’ prior investment in public goods and the like, will face significant financing problems. A single battery plant, solar panel producer or waste management system may find investors and be successful, but for the world to turn green, a lot of unprofitable supportive investment, usually publicly funded, will be necessary.

Uncertainty, the problem at the root of this financing gap, produces market failures: there is a clear demand for money to finance the equally desired green transition, the funds are available, but the two are not connected, because the technological, economic and political uncertainties make it impossible to estimate a reasonable return. As a result, wind farms are not built, EV production and sales stall, and environmental improvement schemes emerge omnly very slowly. Market failures cannot be resolved within the market but require external ‘non-market’ intervention to suspend the potentially debilitating hold-ups. Hence government, or an equivalent private-sector based alternative, need to step in and kickstart investment through industrial policy. The inability of the market to price risk because of the pervasive uncertainty in the green transition leads to severe financing bottlenecks, which seemingly can only be overcome through an external jumpstart.

This uncertainty-induced financing gap is exacerbated by a second crucial problem: current accounting systems handle negative environmental externalities of fossil-based economic activities highly inadequately. Start with the principles of accounting. Because we value future (but certain) gains lower than current losses, we ‘discount the future’ (Doganova 2024) more steeply and therefore are likely to forego necessary and lucrative green investment. Its return is deemed lower and therefore less attractive. While mainly a figment of accounting conventions on net present value, its consequences in the real world are, as Foucault would have been pleased to point out, very real. Since accounting conventions are largely self-referential and self-reinforcing systems, adapting them to think about current brown costs and future green benefits more realistically will be a high mountain to climb. We have a recent example of the difficulty of such a transition to new rules: the simple standardisation of accounting rules that accompanied financial globalisation required significant debate and took many years. In fact, many large companies outside the Anglo-Saxon world, less dependent on the City of London or Wall Street, had two sets of books during the 1990s and 2000s, one for domestic and another for international consumption, and quite a few abandoned Anglo-Saxon capital markets altogether because of the onerous accounting and reporting requirements. Recently, the World Bank and UNCTAD have published papers on new, post-GDP, accounting systems that reflect parallel concerns about the integration of well-being into national accounts.

Not very suprisingly, our existing accounting systems are simply and obviously not very good at handling processes that do not follow standard accounting assumptions. Think of the basic accounting principle of the decreasing value of a good with use: instead of devaluing with use, skills and trust both increase in value with use, and are therefore not easy to incorporate because the line between cost and investment is blurred. The question of unrealised losses poses another conundrum. Today, if you build or do something that has a negative environmental impact, this environmentally detrimental activity counts towards GDP; if you then intervene to rectify it, that also counts towards GDP; the health costs associated with the environmental problem ditto: medical care to treat you after exposure ends up as a plus in GDP. And if you end up in court, your lawyers’ fees also contribute to national income. In short, destruction, reconstruction, mitigation and litigation are all positives from a national account point of view. But they are not, as insurance companies all over the world are beginning to find out: houses in some parts of California have recently become uninsurable because of their exposure to extreme fire risk. They really are costs, and proper accounting would subtract these costs into a net-GDP or -growth figure. Such a model of accounting that recognises costs as negatives on GDP would then take the benefits of not having to handle them into account. But how exactly would that be done, if the costs do not exist because of our own actions to prevent them? Similarly, how do you quantify and monetise a future that involves tipping points, i.e. catastrophic events that you can see coming, and that will destroy all your assets? Because of its conceptual blinders, accounting, the narrative window we use to look at the economic world, is a significant part of the problem.

But assume, for the sake of argument, that we get there, i.e. that accounting systems better reflect the net benefits of a green transition. We then face the problem that calculating all the costs and benefits of the green transition in aggregate is impossible. Many of the costs in particular are likely to be underestimated because they fall outside the conventional frameworks. A simple example is cleaning up industry in Europe by exporting its brown components to China; even if we then were to get China to adopt greener policies, they will likely outsource those to another country, usually in the Global South, which will bear the environmental cost. Or consider the idea that greening industry is likely to be relatively more expensive for poorer countries (more pollution and less money), which will provoke clamours for a compensation package from rich to poor countries (CER 2024). Repatriating parts of the (now greener) supply chains to the core economies – a possibility offered by new, simpler but higher value-added products – will entail dramatic adjustment in the poorer countries where they are currently located: many families will lose their livelihood, with draconian consequences. And since a Tesla (and, presumably other EVs) is estimated to include the same amount of valuable rare materials (lithium, cobalt, etc.) as 10,000 mobile phones (The Conversation 11 Dec 24), the speedy introduction of EVs will impose serious social and environmental costs outside the rich countries that produce and buy them. Parallel to how energy never gets lost in physics, costs never disappear in political economy, but are redistributed between different actors and geographies (Mathei & Hancké 2024). That leads to distributive struggles between the prospective losers and winners of the green transition, and those who are more easily able to mobilise their constituencies – usually not the poor labourer in Africa or southeast Asia – are likely to minimise losses less, increase gains, or both.