Read on Twitter
Thread alert I spent most of last year studying #sustainablefinance. The results are now out in a report for @Academie_be & SFPI-FPIM. Here's the gist (
, 1/n):https://www.academieroyale.be/Academie/documents/Opinio_SFPI_numerique31253.pdf
1) the philosopher's stone of sustainable finance is figuring out how to distinguish sustainable from non-sustainable investments. (2/n)
If you get that distinction right, you can do anything: labels/standards, penalising/supporting factors, subsidies/taxes, mandatory portfolio allocations, even direct public investment, etc. You'd have a guidance system, a GPS, to steer all the other tools. (3/n)
2) To draw this distinction, you obv need a definition of sustainability. But this is not the battle we need to fight right now! @KateRaworth's Donut Econ, the SDGs, etc: that stuff isn't perfect, but it's very very good. I have quibbles, they're in the report, let's move on(4/n)
3) The big challenge, in my analysis, is translating these SYSTEM-level definitions to the INVESTMENT-level. That's the gap in sustainable finance research right now. (5/n)
Good definitions of sustainability give you lots of system-level indicators (CO2 levels, gender inequality, healthcare, etc.). But what's the impact of ANY SINGLE investment on these? How can you separate investments with good system-lvl impacts from those w bad ones? (6/n)
4) My main claim (the 
of the report): you cannot do this by analysing one project at a time. I call this "upwards translation", b/c you're trying to start on the ground (project level) & try to work up to system-level impacts. But upwards translation is impossible. (7/n)
of the report): you cannot do this by analysing one project at a time. I call this "upwards translation", b/c you're trying to start on the ground (project level) & try to work up to system-level impacts. But upwards translation is impossible. (7/n)
5) Why? B/c system-level impacts of any one investment depend on the overall economic system it's embedded in. E.g.: localised agriculture = great if transport & cooling is non-sustainable. Get the farms to the cities, apples from Brabant, not Spanish oranges in Brussels. (8/n)
But if transport & cooling are sustainable (what that looks like is a complicated Q in its own right, just ask @VeraHwe @giulio_mattioli!), you'd want regional specialisation to make best use of different soils/weather patterns etc. V different investment pattern follows (9/n)
Similar: whether investments in electric vehicle production lines & charging points are sustainable or a giant waste of steel, batteries, concrete, engineering time, etc. depends on what happens in urban planning, public transport, remote working, etc. (10/n)
Similar: whether investments in recycling infrastructure are sustainable or not depends on which materials we continue to use at scale (steel? wood? plastics? aluminium? carbon fibre? what will our batteries be made of?) (11/n)
You also see this in life-cycle analyses. Results always depend on methodological choices, esp. scope, + cet. parib. assumption.
Bottom line: System-level impacts of individual projects simply cannot be estimated robustly, certainly not in a dynamic & changing context (12/n).
Bottom line: System-level impacts of individual projects simply cannot be estimated robustly, certainly not in a dynamic & changing context (12/n).
6) If upwards translation can't work, what do we do instead? (13/n)
One option: don't try to draw the distinction in general, focus on quick wins (eg. "dirty list", hello coal), embrace the limits of our knowledge (hello Hayek), use sustainability accounting (put prices & taxes on all the externalities), let the chips fall where they may (14/n)
Lots of problems/debate around sustainability accounting (aka natural capital accounting, full cost acc, true cost acc, etc @theiirc @mwmce) which I cover in the report.
But biggest one: it might move things at the margin, but won't deliver the transition we need (15/n)
But biggest one: it might move things at the margin, but won't deliver the transition we need (15/n)
Why? Systemic transformations are discontinuous. The price mechanism doesn't coordinate them well. (16/n)
This can be hard to see b/c truly systemic transformations are so rare, but consider: lots of market economies, even capitalist ones, existed across history (see e.g. Jairus Banaji's "A Brief History of Commercial Capitalism", as recommended by @njtmulder) (17/n)
Many w/ the technology needed for an industrial revolution, China most famously. But only one (1!) industrialised without a deliberate push: 
(see @hajoonchang's work or Gerschenkron). Those odds aren't good enough. We need the sustainability transition, & we need it now (18/n)
(see @hajoonchang's work or Gerschenkron). Those odds aren't good enough. We need the sustainability transition, & we need it now (18/n)
7) What is to be done, then? Three things, I believe. Together they sum up to "downwards translation", i.e. starting from system-level indicators and ending with investment-level guidance. (19/n)
First: a taxonomy ( @EU_Finance @MartinSpolc). This is the centre. Not b/c it gets things right (see report for critiques), but b/c it's the best tool for turning diverse inputs (climate science, social priorities, supply chain analysis) into investment-lvl classifications (20/n)
Second, a taxonomy is necessarily incomplete & will get things wrong. Hence two supplements: strong system of macro-level indicators to track system-level change (e.g. @EU_Eurostat's https://ec.europa.eu/eurostat/web/sdi/indicators). Are emissions actually falling etc? Taxonomy can't tell you that (21/n)
And mandatory sustainability accounting & taxes, both to track economic efficiency (no better epistemology for that available yet) & to squeeze out gains at the margin.
In light of macro-indicators & sustainability accounting, the taxonomy is then revised regularly (22/n)
In light of macro-indicators & sustainability accounting, the taxonomy is then revised regularly (22/n)
Third, this framework must be drawn up & iterated democratically. The effects will be big & pervasive, the transformation can only succeed if seen as legitimate. French @Conv_Citoyenne is a good example at macro-lvl ( @landemore), bicameral firm at micro-lvl ( @Ferreras_Isa) (23/n)
In sum: it's REALLY HARD to distinguish sustainable from non-sustainable investments. Upwards translation is impossible, starting at project level is a dead-end. Downwards translation may just about work, but it's a colossally difficult endeavour w/ lots of moving pieces. (end)
Report link again here: https://www.academieroyale.be/Academie/documents/Opinio_SFPI_numerique31253.pdf
Big thanks to @Ferreras_Isa, @jvtklooster, @bjmbraun, @DanielZamoraV, @AntonJaegermm, @BringmannKatja, @Lsuttors, @ICassiers, @O_Malay, @martindebrabant, @BrunoColmant, @DidierViviers for support, discussions & feedback!
Big thanks to @Ferreras_Isa, @jvtklooster, @bjmbraun, @DanielZamoraV, @AntonJaegermm, @BringmannKatja, @Lsuttors, @ICassiers, @O_Malay, @martindebrabant, @BrunoColmant, @DidierViviers for support, discussions & feedback!
Might be of interest to @Fridays4future @econ4future @adam_tooze @MESandbu @LaszloAndorEU @NEF @alfie_stirling @Frank_vanlerven @MehreenKhn @attractamooney @alybatt @KateAronoff @triofrancos @BuddyYakov @abenanav @onegoodchart @IMateosylago @DanielaGabor
