Climate levers: sufficiency, technology or policy?

Between 2005 and 2015, eighteen developed countries saw their CO2 emissions fall. When researchers decomposed those declines, they identified two main drivers: the substitution of fossil fuels by renewables, and reduced energy consumption, itself partly attributed to slower economic growth. No significant contribution was attributed to the promotion of sufficiency, meaning campaigns aimed at convincing citizens to consume less out of ecological conviction.[1]

Romain Barbe

Founder · Mieux Donner · Reading time: 18 min

This finding is unsettling for anyone personally engaged in reducing their carbon footprint. It raises a harder question for anyone who wants to donate effectively to climate charities: if motivated sufficiency has not driven the observed reductions so far, should we still fund it? And if not, where should we put our money?

This article offers a framework. It first distinguishes physical levers (what actually reduces CO2) from action instruments (how we act on them). It then looks at what the data says about the observed reductions and what really produced them. It closes on the place of individual action and on why leading climate philanthropy evaluators recommend almost exclusively policy advocacy and support for innovation, rather than individual awareness campaigns.

Physical levers and action instruments

Before looking at what works, we need to be clear about what we are looking at. Climate debates often blur two dimensions of different nature: what physically reduces emissions, and how we go about triggering change. A quick comparison makes the distinction clear. If you want to lose weight, you can eat less or move more (those are physical levers), but you can go about it by following a diet, seeing a doctor, or joining a gym (those are instruments).

The five physical levers

At the physical level, five broad families of levers can reduce greenhouse gas emissions. This classification aligns with the framework used by the IPCC in its sixth assessment report, which combines the demand-side categories "Avoid, Shift, Improve".[2]

The five action instruments

Action instruments are the concrete ways to activate physical levers. The IPCC devotes chapters 13 and 16 of its sixth assessment report to public policies and technological innovation. The classification below synthesises those approaches into a grid useful for evaluating a philanthropic strategy. Five families of instruments structure most climate action.

This distinction between levers and instruments matters because the same charity can act on several intersections. An organisation advocating for a carbon tax acts through the "policy advocacy" instrument but affects decarbonised energy, efficiency and sufficiency simultaneously (since emitting anything becomes more expensive). Conversely, a charity raising awareness about food sufficiency activates the "norm change" instrument on the single "sufficiency" lever. When evaluating a climate philanthropic strategy, we should ask which lever it targets and how large that lever is, but also through which instrument it acts, because the instrument also determines the leverage of a donated euro.

What actually drove emissions down, or slowed them

Global CO2 emissions from fossil fuels have not peaked. According to Our World in Data, drawing on Global Carbon Project data, they kept rising in 2024.[5] The right question at the planetary scale is therefore not "what drove emissions down?" but "what kept them from rising further?" And in the countries that actually did reduce emissions, the question becomes: how did they do it? The answers converge on the same levers, and the promotion of sufficiency is not one of them.

The decomposition of reductions in 18 developed economies

The reference study on this question was published in 2019 in Nature Climate Change by Corinne Le Quéré and her team. It decomposed the emissions declines of eighteen developed economies between 2005 and 2015. The conclusion is clear: those declines are explained by two main drivers, the substitution of fossil fuels by renewables and the drop in energy consumption, the latter partly attributed to slower economic growth.[1]

The authors conclude that policies supporting renewables and energy efficiency did contribute to the observed reductions. But they argue that significant future cuts will require policies that directly penalise carbon emissions. No distinct contribution is attributed to sufficiency promotion. This does not mean that no behavioural change took place: sufficiency simply does not appear as an autonomous explanatory factor in the decomposition.

Electricity decarbonisation as the main driver

The lever that emerges from every analysis in countries that reduced their emissions is the transformation of the electricity mix. The UK is a striking case. According to Our World in Data, the country produced more than half of its electricity from coal at the end of the 1980s. That share is now down to a few percent, and British electricity generation relies largely on gas, renewables and nuclear. Over the same period, UK CO2 emissions per capita fell sharply while GDP per capita kept rising.[6] Electricity decarbonisation is what allowed this decoupling, not a collective sufficiency effort by the British.

According to Our World in Data, other countries have achieved similar decoupling — even when accounting for emissions embedded in imported goods — including France, Germany, Sweden, Denmark, Italy, and other developed economies. In every case, the dominant lever is energy decarbonisation, complemented by energy efficiency.

At the global level, renewable growth is real but remains far from sufficient given the global growth in energy demand. According to Our World in Data,[7] renewables (including hydro, wind, solar and other sources) now account for about 30% of global electricity production, against 9% for nuclear and the rest for fossil fuels. But the total energy mix, which includes the direct combustion of fossil fuels for industry, transport and heating, remains dominated at around four-fifths by coal, oil and gas.

Energy efficiency: a quiet but massive lever

The second driver of the observed reductions is energy efficiency. According to the International Energy Agency, energy efficiency improvements avoided about seven gigatonnes of CO2 cumulatively between 2010 and 2022, a very significant volume when compared to annual global emissions.[8] The IEA also estimates that the five key low-carbon technologies (solar PV, wind, nuclear, electric vehicles, heat pumps) have avoided about 2.6 gigatonnes of CO2 per year since 2019.[9] Without them, the rise in global emissions would have been three times greater.

European gas after 2022: imposed sufficiency, not motivated sufficiency

The European energy crisis of 2022 offers the clearest example of a rapid drop in energy consumption attributable to behavioural reduction. Combined gas and electricity consumption in France fell by about 12% between August 2023 and August 2024 compared to the 2018-2019 baseline, according to the official assessment of the sufficiency plan.[10] Across Europe, natural gas consumption has dropped by more than 20% since 2021.

Much of that decline, however, was not due to environmental education. As Nicolas Goldberg of the Terra Nova think tank points out, those reductions also stem from price shields and from the surge in energy prices following the invasion of Ukraine.[11] The provisional 2024 energy balance published by the French Ministry of Ecological Transition confirms that the initial efforts are eroding: France's final energy consumption was roughly stable in 2024, up 0.3% once corrected for climate variations.[12] When prices come back down, a share of the "sufficiency" evaporates.

Meta-analyses on awareness campaigns

The question then arises of interventions explicitly designed to promote low-carbon behaviours. A meta-analysis published in 2019 in Nature Communications by Nisa and colleagues synthesised the results of 83 randomised controlled trials covering more than 3 million observations. The authors' conclusion is severe: behavioural interventions promote climate mitigation only to a small extent during the intervention itself, and the authors found no evidence of lasting effects once the intervention ended.[13] The authors note that behavioural plasticity is low in most domains (with the notable exception of recycling).

A reanalysis published in 2020 by Bergquist and colleagues in the same journal reached a slightly more favourable estimate (average effect of around 0.2), but the authors of both studies converge on the point that purely behavioural interventions are strongly amplified when combined with structural change (prices, infrastructure, mandatory norms).[14] In other words, what works is less the message than the context around it. A sufficiency campaign in a canteen where the plant-based option is set as default has a much larger effect than a campaign where it remains an active choice.

This literature does not say that awareness is useless. It says that its effectiveness, isolated from other levers, is weak and not durable. For the philanthropic reasoning that follows, this is a decisive point.

Motivated sufficiency and imposed sufficiency

Not everything that looks like sufficiency is the product of an ecological choice. This distinction is essential for understanding the reductions observed in Europe and what they mean for climate philanthropy. When a Parisian gives up their car, it is almost never because a charity convinced them: it is because the city has made driving difficult (low-emission zones, paid parking, dedicated cycle lanes) or because running costs have become prohibitive. When a German household turned down the heating in 2022, it was not because they had read an IPCC report: it was because the gas bill had doubled.

Two broad categories of low-carbon behaviour can be distinguished in society. Motivated sufficiency is adopted by an individual because they have internalised the climate stakes. It is what sufficiency-promoting charities aim at: talks, workshops, climate fresks, campaigns, educational programmes. It requires work on representations, values and ecological awareness. Imposed or context-driven sufficiency refers to low-carbon behaviours that result from external constraints: high prices, infrastructure that makes the carbon-heavy option inaccessible, regulations that limit or ban it.

The examples covered in the previous section (drop in European gas consumption after 2022, French sufficiency plan, peak car in large European cities) fall largely in the second category. This does not disqualify them as contributions to lower emissions: they count in the statistics. But it radically changes what we can conclude about the effectiveness of sufficiency promotion as a philanthropic strategy. What worked was less pedagogy than prices and rules. And prices and rules are the product of political work, not of an awareness campaign.

Flygskam, or the gap between culture and emissions

The Swedish flygskam movement ("flight shame") is the best-known example of large-scale motivated sufficiency in a single country. Born in 2018 around the figure of Greta Thunberg and popularised by the media, it triggered a cultural shift in the relationship to air travel. The Swedish airport operator Swedavia documented a notable drop in domestic passenger numbers, and several surveys confirmed that a growing share of Swedes reported changing their travel habits for climate reasons.[15]

The overall picture is more ambiguous, however. As Jörgen Larsson, a researcher at Chalmers University and one of the specialists on the topic, sums up, emissions per passenger-kilometre in aviation fell by around 2% per year, but the total number of passengers kept growing by more than 3% per year before Covid. Overall, aviation emissions kept rising, even in the country where motivated flight sufficiency is culturally most advanced.

This does not diminish the value of the individual choice to fly less. It raises a quantitative question: if motivated sufficiency was not enough to lower aviation emissions in Europe's most convinced country, by what mechanism would a French awareness campaign do better? The question is open, and the burden of proof lies with those who propose this kind of philanthropic intervention.

Personal sufficiency retains several justifications independent of its direct macro impact: it preserves ethical consistency between the values we defend and the choices we make, it plays a social-signalling role that influences those around us and spreads norms, and it steers markets by lowering the relative cost of low-carbon options. One caveat matters, though: several studies in moral psychology document a risk called single action bias or moral licensing, whereby an action perceived as morally positive can unconsciously reduce engagement in other areas, such as support for binding public policies.[16] It is therefore better to combine personal sufficiency, civic engagement, and giving to effective organisations.

The place of individual action within the system

It might seem like individual action and systemic change are at odds. Several scientific voices reject this dichotomy, and one of the clearest is Hannah Ritchie, a data scientist at Oxford University and lead researcher at Our World in Data. Her book Not the End of the World (2024) and her regular writings on the Sustainability by Numbers blog offer a more nuanced way of thinking about the relationship between individual behaviour and collective transformation.[17]

Why opposing individual and systemic is a false dichotomy

According to Ritchie, the question "should we act individually or collectively?" is poorly framed. The two levels are not opposed but do not have the same mechanism of action. Individual action works mainly through three channels: it directly (but marginally) reduces personal emissions, it sends signals that influence markets and norms, and it prepares the ground for political change that voters would otherwise be less inclined to support. Collective action works through the evolution of rules, infrastructure and prices, which simultaneously change the choices of millions of people.

The leverage of an individual action is not limited to its direct effect. A person who cuts their footprint from 9 to 4 tonnes per year has a cumulative effect over thirty years of around 150 tonnes. That same person, through their vote, their influence on those around them, their career choices, or a donation to an effective policy-advocacy organisation, can contribute to emissions reductions several orders of magnitude larger.

This logic aligns with what Mieux Donner regularly writes based on Ritchie's own data: following the three most impactful individual actions (giving up a car, adopting a plant-based diet, avoiding transatlantic flights) saves about 6.2 tonnes of CO2 equivalent per year. These are excellent initiatives, but at the scale of the climate crisis, they are not enough. The lever that unlocks scale lies elsewhere.

Personal actions as signals

Part of the value of individual action is not measured in avoided personal CO2 tonnes, but in the signals it sends. This is one of Ritchie's central points: our consumption, mobility and food choices help structure the markets and social norms that in turn make ambitious public policies possible or impossible.

Three mechanisms are documented. The first is the market signal: buying a low-carbon product, choosing the train, opting for a plant-based dish at a restaurant lowers the relative cost of these options through scale effects and steers industry innovation. An electric car bought today makes the next one cheaper for someone else. The second is the diffusion of social norms: behaviours spread through concrete social networks (families, colleagues, neighbourhoods) before they diffuse through the media. A vegetarian person in a group of friends statistically shifts the choices of the others. The third is support for public policy: residents of a city where cycling has become visible more easily accept cycle lanes, and voters who have adopted a heat pump become less hostile to gas standards.

Ritchie also stresses an important point for the psychology of climate action: a majority of people in every country support climate action, but the public massively underestimates that support. Moving past the idea that "no one cares about the climate" is itself a political act, because this false perception fuels passivity.

Carbon footprint and action footprint

Ritchie proposes to distinguish two measures of an individual's climate action.

The carbon footprint measures the greenhouse gas emissions directly and indirectly attributable to a way of life:

• food
• transport
• housing
• goods consumed

This is the measure popularised by online calculators, and it is what is at stake when we talk about personal sufficiency.

The action footprint (sometimes called "carbon handprint") measures a person's capacity to influence the emissions of others and of the system:

• voting
• volunteering
• career choice
• influence on those around us
• civic engagement
• donations to effective organisations

Unlike the carbon footprint, which is bounded by our standard of living (you cannot get to zero while living in France), the action footprint is only bounded by our capacity to steer resources and decisions.

This distinction is useful for philanthropic reasoning. Cutting your personal footprint by 5 tonnes per year has real value, limited to those 5 tonnes. Giving the equivalent budget to an effective policy-advocacy or low-carbon R&D organisation can, according to evaluators' estimates, have an effect one order of magnitude larger. Personal sufficiency ensures consistency and signal; a well-targeted donation ensures leverage.

It is on this leverage that the recommendations of climate philanthropy evaluators rest. Let us look at how they identify it.

The criterion of marginal impact

Once the physical levers and action instruments are identified, how do the specialised climate philanthropy evaluators translate all this into concrete recommendations? The answer lies in the notion of marginal impact, which structures the evaluations of Giving Green, Founders Pledge and Mieux Donner.

Marginal impact and average impact

The marginal impact of a donation is the additional effect it produces, beyond what would have happened without it. It is measured through counterfactual reasoning: what would have happened with this money, or without it, in any case? This question is fundamental for Mieux Donner, and it leads to very different conclusions than the average impact of a cause would.

Take an example. Electricity decarbonisation is, as we saw, the top physical lever behind observed reductions. That does not mean we should give massively to solar deployment projects in France: the sector is already backed by tens of billions of euros in public funding, by mature markets, by profitable private operators. The philanthropic euro added to a rooftop solar project in Lyon changes almost nothing compared to what would have happened. Its marginal impact is low, even if the average impact of the lever is high.

Conversely, some building blocks of the climate transition are structurally underfunded: methane regulation in the oil and gas industry, R&D on deep geothermal or meat alternatives, advocacy for industrial standards in emerging countries. These are levers where each additional euro can accelerate something that would otherwise drag on for years. Marginal impact there is very high, even if the average aggregate impact of the cause at the global level is harder to estimate.

This logic explains much of the evaluators' recommendations: not the most important lever in absolute terms, but the lever where an extra euro makes the biggest difference.

Scale, tractability, neglectedness

To evaluate this marginal impact, Mieux Donner uses a three-criteria framework derived from the ITN model (Importance, Tractability, Neglectedness) used by international evaluators, which we translate as scale, tractability and neglectedness.

• Scale: the size of the problem, meaning the number of people or ecosystems affected and the intensity of the impact. For the climate, the scale is enormous: a global, intergenerational problem affecting health, agriculture, the habitability of entire regions. On this criterion, all climate sub-causes score very high.
• Tractability: the real capacity to act, meaning the existence of documented solutions, organisations able to deploy them, identifiable levers. This is where the first hierarchy among levers appears. Electricity decarbonisation has very high tractability (technologies exist, policies have demonstrated effects). Motivated sufficiency has tractability that is harder to assess, as shown by the meta-analyses cited above.
• Neglectedness: how well the cause is already funded by other actors (governments, markets, foundations). This is where most of the final selection is decided. Mature European solar PV is well funded. Methane capture in oil and gas is much less so, despite its potential. Deep geothermal even less. This criterion explains why evaluators steer donations toward organisations working on technically powerful but structurally underfunded levers.

Why policy advocacy comes out on top

When these three criteria are applied to the full set of evaluated climate organisations, advocacy for ambitious regulation consistently comes out on top, followed by support for innovation on underfunded critical technologies. Several reasons explain this.

First, institutional leverage is unmatched by other instruments. A single regulation can, within a few years, change the choices of millions of actors. The European methane regulation adopted in 2024, which Clean Air Task Force (CATF) worked on closely with the European institutions, should cut methane emissions by a volume equivalent to the total annual emissions of Germany.[18] No awareness campaign reaches a comparable order of magnitude.

Second, these organisations operate in neglected niches. CATF has repeatedly targeted technically powerful solutions that no one else was defending politically: renewing gas infrastructure to limit methane leaks, supporting deep geothermal, the 45Q carbon capture tax credit in the US. This niche logic is exactly what the neglectedness criterion rewards.

Third, uncertainty is integrated into the reasoning. As Mieux Donner explains in its methodological article on carbon impact calculation, it is not possible to attribute a precise number of avoided CO2 tonnes to advocacy work. Founders Pledge, which produced the most rigorous available evaluation of CATF, estimated a cost of about one dollar per tonne of CO2 avoided, noting that this is a prudent estimate. The report even points out that the real effectiveness could be ten or a hundred times higher. This estimate is now shared by Giving Green and by Effektiv Spenden.[19]

This order of magnitude, which gives the formula about €1 donated = about 1 tonne of CO2 avoided, needs to be handled with great care.

The absence of sufficiency in current recommendations

At the time of writing, Giving Green's 2025-2026 climate recommendations include Clean Air Task Force, Industrious Labs (industrial decarbonisation), Good Food Institute (meat alternatives), Project InnerSpace (geothermal), Future Cleantech Architects (R&D on critical technologies), and Opportunity Green (advocacy for developing countries). No organisation focused on promoting individual sufficiency appears among these recommendations. The climate fund operated by Founders Pledge has a similar composition, oriented toward advocacy and innovation.

Evaluators do not judge sufficiency to be without value. At the current stage of their research, they have not identified any organisation whose sufficiency-promotion work simultaneously meets the three criteria, mainly because environmental awareness is not an underfunded area in France or Europe (unlike technical advocacy on methane or R&D on geothermal) and because the marginal impact of a specific charity on the diffusion of low-carbon behaviours is weak. This could change: if in the coming years an organisation shows indications that its marginal impact is strong and more cost-effective, it could appear in the recommendations.

Combining personal sufficiency with climate giving

What drove CO2 emissions down in countries that decarbonised, or slowed their rise elsewhere, is energy decarbonisation, energy efficiency, and in some cases regulation and prices. Sufficiency motivated by an ecological choice retains value for ethical consistency, for the diffusion of social norms, and for the signal it sends to markets and decision-makers. But it has not so far been a quantifiable driver of the observed reductions. Evidence-based climate philanthropy evaluators therefore orient their recommendations toward ambitious policy advocacy and support for underfunded critical technologies, rather than toward sufficiency awareness. This orientation comes from counterfactual reasoning and the marginal impact criterion. It does not pass judgement on the value of individual lifestyles.

Frequently asked questions