Comment to “A Proposal for a Carbon Wealth Tax: Modelling, Empirics, and Policy” (DOI https://doi.org/10.1515/jbnst-2024-0078) by Willi Semmler and José Pedro Bastos Neves

1 Background

The existential risks of climate change are becoming ever more undeniable – notwithstanding the broad-based rolling-back on mitigating measures to be expected from the new U.S. Administration. Economists’ preferred tool to address the “greatest externality” Nicholas Stern is to put a price tag on green-house gas (in particular, CO2) emissions. This could be via a price (i.e. taxes, accounting for the damage resulting from the externality), or by way of quantitative ceilings (emission allowances). Of course, in a “second- or third-best world”, departures from the “single price of carbon in all places, dates, and uses” (J. E. Stiglitz 2019) are well-advised. These include subsidies, standards, regulations etc. – often in the perspective of “directed technological change” or “green industrial policy” (Rodrik 2014).

While the diagnosis of potentially catastrophic climate change is uncontentious, the ways and means to address its consequences are (highly) disputed. One particularly pernicious issue relates to the distributional consequences of mitigation measures. Regularly, they are regressive, i.e. they amplify prevailing tendencies at increasing inequality, income as well as wealth. (A recent example highlighting this conflict, has been the protests of the French ‘yellow vests’ movement. The seemingly marginal increase in a levy on gas fanning particularly the resistance of employees in rural areas. To get to their jobs, these people are dependent on using their cars. In many polities, the often latent resistance to implement climate-friendly policies meanwhile manifests itself in the increasing approval rates of political programs promising to turn back the clock. In fact, this is not only an electoral issue – difficult to disregard by political parties –, but it also concerns initiatives to ‘green’ finance. In the U.S., for instance, 12 states are threatening asset managers pursuing ‘woke’ agendas with litigation (Financial Times, January 7). BlackRock (as well as numerous other financial institutions) just left the net zero asset managers group (Financial Times, January 9). Others never became members.

Against this background, an instrument that promises a solution to both – mitigating the consequences of climate change and concurrently reducing wealth inequality – is, obviously, highly attractive. That is what Pedro Bastos and Willi Semmler’s carbon-based wealth tax proposal promises to deliver. In Section 2 we will sketch the core elements of the proposal. Section 3 highlights assumptions, some of them implicit. Section 4 concludes with proposals to enhance the probability of implementation.

2 The Bastos-Semmler Proposal

To reduce the further accumulation of green-house gases and contain global warming, Bastos and Semmler suggest a carbon-based wealth tax (CWT). That is, the tax base should be wealth arising from the ownership of ‘brown’ assets. Holders of green assets would be exempt. (The alternative proposal, based on the polluter pay’s as well as the ability to pay principle, suggests taxing the wealthiest [the Billionaires], regardless of the sources of their wealth – whether brown or green; see, e.g. (Saez and Zucman 2019; Saez and Zucman 2023)).^[1]

With reference to Musgrave (Musgrave 1959) the justification is that wealthy owners of brown assets are responsible for substantially higher share of “public bads” (= negative externalities, unaccounted for in prices).

The CWT would be assessed on the normal rate of return arising from the ownership of brown assets, i.e. a tax on income derived from an asset (not the asset value as such). It would not be appraised on the stock of brown capital. This approach, B-S argue, also mitigates implementation problems. A CWT would be easier to collect since less impacted by information (valuation) issues (including evaluation of net worth). Therefore, tax avoidance (underreporting etc.) or evasion would be more difficult. Otherwise exploitable loopholes would be shut. B-S justify their proposal with reference to Richard Musgrave’s proportionality principle: Given that owners of brown assets are to a large degree responsible for the creation of public bads, they should also pay a commensurate tax. This should come with allocative consequences, more impactful than flow-oriented conventional carbon taxes, targeting CO2 emissions resulting from consumption or production. What concerns incidence – the use of brown assets is a derived demand – B-S observe (B-S 2021) that a CWT would be largely paid by the brown sector. Of course, as B-S state, the identification of what amounts to a brown asset is of the essence (see Section 3).

B-S (B-S 2021) derive their propositions elegantly from a dynamic portfolio-perspective in the Merton tradition, allowing for time-varying returns (except for the safe asset – which, however, with 3 % p.a., is supposed to deliver a rather high return). The share of the brown asset is the upshot of the (representative) investor’s utility maximizing wealth allocation. Returns changing over time are impacted by (1) a product (excise) tax, (2) a wealth (brown capital) tax and (3) green subsidies. Subsequently, two cases are evaluated: (a) a brown capital tax only and (b) the use of CWT revenues to subsidize green investments.

Simulating the model against data produces three results: Taxing carbon emission at the level of products (excise tax) does not have much of an effect on wealth distribution. (Wealth holders largely compensate the tax in a Barro-Ricardian way.) Therefore, an excise tax also does not change investment patterns. This is, however, what a CWT, by taxing returns, promises to achieve, although only when tax rates are substantial (40 %). Moreover, the simulations also show a trajectory over time for the share of the green asset, being larger in an initial phase, augmented in the case of a higher tax rate. Since green subsidies positively impact returns, they also amplify wealth accumulation dynamics, i.e. lead to more green investments.

3 Queries for the Bastos-Semmler Proposal

Brown capital is a private asset and a societal liability. Therefore, changing incentives at the source – by taxing returns from making use of these assets – should change wealth accumulation dynamics, the patterns of green vs. brown investments. In the following, three points will be raised.

First, incidence. While B-S claim that it will be largely brown asset owners who pay, this is not evident. But it is important for their argument according to which their approach kills two birds with one CWT stone – the allocation and the redistribution one. The use of brown assets is a derived demand, contingent on the final products households buy. Therefore, in a partial equilibrium perspective, incidence depends on relative demand and supply elasticities.

According to B-S, given the trajectory of expected returns of green and brown assets, it is the optimizing behavior of investors which determines their respective share in wealth accumulation. While the behavior of (institutional) investors is evidently impactful, other financial intermediaries obviously also do play a role. As do, at firm level, internally generated funds. As far as returns in capital markets are a benchmark for decisions of other agents in the financial system (banks, private equity firms and ‘real’ firms and their decisions about their cash-flows) this only ‘dilutes’ the argument of B-S. But, depending on respective circumstances, this could be a relevant impact. And, of course, one can be less impressed by the “functional efficiency” (Tobin 1984) of capital markets (or financial intermediaries) to discern the welfare improving trajectory to achieving the net-zero transition.

The third point is closely related. It concerns identifying brown vs. green assets. This could be physics – derived from the CO2 content of using fossil fuels, hence science-based and uncontroversial. But it is not.

There are several regulatory efforts to define brown assets. And all rely on disclosure obligations and efficiency of markets in processing this information. The U.S. Security and Exchange Commission, for instance, “require(s) a registrant to disclose, among other things: material climate-related risks; activities to mitigate or adapt to such risks”. To produce accountability, the SEC also requires “…information about the registrant’s board of directors’, oversight of climate-related risks and management’s role in managing material climate-related risks… (as well as) information on any climate-related targets or goals that are material to the registrant’s business, results of operations, or financial condition.”

As concerns the EU, the Corporate Social Responsibility Directive (CSRD 2022), which entered into force in January 2023 and will be fully applicable from 2028 onwards, defines sustainability reporting requirements. The CSRD obliges to report effects and feedback beyond the reporting firm (“double materiality”).

Moreover, the EU-Taxonomy provides a classification system of economic activities aligned with net zero trajectory by 2050, i.e. mitigation and adaption to climate change mitigation (including Scope 1, 2 and 3 GHG emissions), water and marine resources, resource use and circular economy, pollution, and biodiversity and ecosystems.

Moreover, further initiatives have been launched. This includes, inter alia, the Network for the Greening of the Financial System (with central banks and supervisors as members). And in the case of the ECB, the Single Supervisor Mechanism has conducted in 2021 an economy-wide climate stress test. In 2023, the SSM defined “Good practices for climate-related and environmental risk management”.

However, as the debate about the famous EU taxonomy has made patently clear, conflicting interests are at play. This often leads public sector authorities to delegate decisions to apparently disinterested gatekeepers. However, delegating decisions on the quality of firms’ environmental conduct to rating agencies seems to produce substantially more problems than in the assessment of default probabilities (Kotz and Schäfer 2013). Their evaluations go all over the place, as documented in an appropriately titled article by Berg, Köbel and Rigobon (Berg, Köbel, and Rigobon 2022): “Aggregate confusion”. In fact, a recent updating of the project by Roberto Rigobon and Florian Berg produced even more problematic results. What concerns the impact of policy interventions, the use of indices blending environmental, social and governance scores is certainly part of the problem. ESG aggregates incommensurable, multi-criteria dimensions. “E” alone is more encompassing the climate risk. Unfortunately, making use of supervisors’ templates to measure environmental risks is not – yet? – more promising. As documented, based on credit-level data, Sastry, Verner and Ibanez (Sastry, Verner, and Ibanez 2024) show that net-zero promises of banks were not detectable neither in divestments (from brown industries) nor in new green investments. Again, the title of the paper is telling: “Business as usual”.

4 Concluding Remarks

To be clear, while, on a normative basis, I completely share the allocational and re-distributional objectives Bastos and Semmler pursue, my issues concern the binary delineation between brown and green assets (based on the assessment of rating agencies) as well as feasibility, i.e. implementation. To address the first issue, thinking about independent expert committees (not unlike fiscal boards) would be worthwhile. The second problem is substantially more difficult. What would be the political conditions that would allow governments to introduce a CWT on domestic owners of fossil resources?

Moreover, the apparently most effective approach (i.e. combining a CWT with green subsidies) would be up against the typical problems of earmarked revenues, underwriting that these revenues are effectively deployed to the dedicated purposes: Money is fungible, as the saying goes. Still, combining a CWT with directed technological change (green industrial policy, subsidies for accumulation of green assets) (Semmler and Nyambuu 2023) (ch. 8 and 9), seems appropriate. It could allow addressing the wedge between private and social returns (substantial positive externalities, high uncertainty, mispricing of GHG emissions, etc.).

Given the second- or third-best environment in which the mitigation of climate risks has to be implemented, one has to ponder how a CWT could be aligned with other policies or instruments (in limiting the increase in global warming and achieving net zero. Hence, the question is: How does CWT fit into carbon pricing (EU ETS), taxes, subsidies for green investments, standard setting, regulations (banking, capital markets)? (J. Stiglitz, Stern, and Taylor 2023; Pisani-Ferry and Mahfouz 2023).

Finally, addressing the risks of climate change is about producing a global collective good. With the scope for international cooperation shrinking, important players unwilling to cooperate, the issue of leakage (and carbon border adjustment mechanisms) looms ever larger. With the new U.S. Administration pursuing a rolling-back policy, at times completely reversing previous initiatives, even the implementation of current mitigation frameworks has now become deeply endangered. The European Commission as well as important EU member states are responding to the new U.S. approach by “simplifying” and “postponing” previously agreed upon regulations as well as by limiting their scope (see, e.g., European Parliament 2025). Of course, this does not speak against the very interesting proposal of Bastos and Semmler, a proposal at a technical (or technocratic) level. But, very unfortunately, at a higher plane, what one meanwhile calls geopolitics, hurdles have been rising substantially of late.

However, to end on a less skeptical note, recently, there have also been positive cases of international cooperation – automatic exchange of bank information between tax authorities; OECD Pillar 2: minimum tax rate on profits (Zucman 2023). Thus, at some point, conditions for well-conceived cooperative approaches might re-emerge. Yet, as the great Dick Cooper has demonstrated, it can take much time for nations to agree (Cooper 2001). But time is running out.