The Limits to Growth and the Return of Models of Doom

How a disputed computer model became one of the most influential ideas in modern environmentalism

A recent op-ed published in The Guardian took the view that economic growth was a ‘doomed strategy’. Signed by Tomas Piketty and Joseph Stiglitz alongside degrowth advocates Jason Hickel and Kate Raworth, it argued that a growth-centred economic model had reached its limits and that poverty reduction should no longer depend on economic growth. Yet this diagnosis is far from universally accepted. Philippe Aghion and Simon Johnson have recently argued that Europe’s relative economic decline stems not from excessive growth but from insufficient dynamism, innovation, and creative destruction.

This is an old debate, fifty-four years ago, the authors of The Limits to Growth asked whether an economic system organised around continuous material expansion could persist indefinitely on a finite planet. Their answer was controversial enough to provoke decades of criticism. The debate has changed over time and the question is no longer whether there are limits to growth, but what those limits mean, how they should be understood, and whether Meadows and colleagues were asking the right question in the first place.

The most glaring inconsistency of the LTG model (World3) is that if it is run from 1600 or 1800 it predicts collapse roughly a century later, a point first highlighted by Kevin Kelly. The world did not collapse in 1700 or 1900. This is a simple test of forecasting a model’s internal consistency, yet neither the original research team nor later updates subjected World3 to it. The model would also fail to generate the Industrial Revolution and, as Dana Meadows acknowledged, could not ‘take the world from the Industrial Revolution to whatever follows next beyond that’.

The World3 model was intended to show how an economic system might overshoot ecological limits and collapse. Yet, despite its weaknesses, the model continues to shape contemporary debates about degrowth, doughnut economics, and planetary boundaries. The reason LTG survives despite decades of criticism is that it stopped being judged primarily as a forecasting model and instead came to be valued as a moral narrative. Its influence rests less on the accuracy of its predictions than on the worldview it offered: a world of limits, overshoots, and ecological constraint.

The Club of Rome’s Limits to Growth (LTG, published in March 1972), and its 1974 follow-up volume responding to various critics, have had an extraordinary intellectual influence. Its ideas continue to shape contemporary environmental thought, most notably through the Planetary Boundaries framework, which explicitly ‘builds on and extends approaches based on limits-to-growth’. The LTG continues to be cited within academic scholarship. For example, a recent degrowth article published in the Lancet Planetary Health includes the scenarios of the Limits to Growth without any commentary on the external and internal validity of the underlying computer model.

The influence of the Limits to Growth

The LTG argued that there were limits to economic growth and that the world needed to change tack or risk ‘overshoot and collapse’. As the Club of Rome marked the 50^th anniversary of the publication of the Limits to Growth in 2023, it hosted a webinar (The Limits to Growth + 50) with Dennis Meadows, lead author of the LTG, Johan Rockström, the lead author of the Planetary Boundaries reports, and Jayati Ghosh, a development economist and a commissioner associated with the Club of Rome’s Earth4all. After Meadows presented on the LTG, Rockström lauded the prescience of Dennis Meadows and the LTG project:

Dennis meadows you are absolutely one of my top heroes in the world and you’ve influenced me and so many of my peers in the world. I would say you’re the father behind system dynamics modelling, what has become modern earth system modelling and you’re so humble when you always emphasize that you were not doing any predictions but you were sitting there with your team in the in the early 1970s... So it has really guided science and our understanding of human relations with your system for this whole time...We still follow a pathway to disaster

The LTG utilised the system dynamics framework developed by MIT management professor Jay Forrester. The model (World2) was outlined in his 1971 book World Dynamics. The LTG was the result of an MIT based project which updated and applied a version of Forrester’s model (World3) to the global economy. The intellectual underpinnings of World Dynamics were Malthusian as Forrester outlined:

Ever since Malthus stated his propositions relating population and food some 150 years ago, the validity of his assumption that food imposes an ultimate limit on population has been debated. The continued growth of population and the rise in the productivity of agriculture are often cited to refute Malthus. But it is undeniable that Malthus stated one ultimate barrier to unending population expansion. His assertion is not erroneous; it is merely incomplete.

The incompletion, as far as Forrester was concerned, was that Malthus had not specified other constraints on population growth. The World3 model was based on a series of equations that linked population, capital investment, natural resources, pollution, and food production. Effectively it is an attempt to model the interrelation between each component. The starting point was that growth in each of the variables had been exponential up to that point and that exponential growth could not be expected to continue, Forrester sought to determine what would bring the fall in growth and what will be the ‘barriers when growth goes too far’. Forrester’s model effectively predicted a collapse in world population in the early twentieth-first century.

The work of the LTG was undertaken by an interdisciplinary research team but with a core expertise in system dynamics. The report was provocative with a central message which identified an ‘overshoot and collapse’ of the global economy. LTG was a non-technical summary of a computer model of the global economy, the report focused on an assumed exponential growth in population, industrialisation, and pollution, and its impact on food production, and consequential resource depletion more generally. In conclusion, it warned of a collapse of the global economic system within 100 years (see Figure 1). The LTG concluded that:

1. If the present growth trends in world population, industrialization, pollution, food production, and resource depletion continue unchanged, the limits to growth on this planet will be reached sometime within the next one hundred years. The most probable result will be a rather sudden and uncontrollable decline in both population and industrial capacity.

2. It is possible to alter these growth trends and to establish a condition of ecological and economic stability that is sustainable far into the future. The state of global equilibrium could be designed so that the basic material needs of each person on earth are satisfied and each person has an equal opportunity to realize his individual human potential.

3. If the world’s people decide to strive for this second outcome rather than the first, the sooner they begin working to attain it, the greater will be their chances of success.

Figure 1: World model standard run

A central focus of LTG was on depletion of non-renewable resources (such as aluminium, chromium, coal, cobalt, iron, lead, manganese and mercury), which would precipitate this global collapse. In one scenario, available global resources were assumed to double but even this did not prevent a collapse (see Figure 2).

Figure 2: World model Resources Doubled

The LTG was so widely read that a contemporary cover of Newsweek ran with the headline ‘Running out of everything’ next to a picture of Uncle Sam looking into an empty cornucopia. The report also coincided with the oil crisis of October 1973 to March 1974, which saw oil prices quadruple (see Figure 3). This seemed to support the gloomy prediction that humanity was running out of resources.

Figure 3: Index of crude oil prices in 2000 dollars (1990-2000 = 100)

Models of Doom

While the modern legacy of the LTG has been an uncritical acceptance of the warnings, contemporaries were less convinced. One of the most stinging rebuttals of the entire approach of the Limits to Growth came from the Models of Doom, a book written by an interdisciplinary research team that looked at various components of the LTG.¹

Some of the main criticisms related to the data, or rather the lack of data. Many of the model parameters were inferred based on guesswork. The old aphorism ‘garbage in, garbage out’ was quite applicable in the case of the LTG; given the Malthusian preferences of the team, it is perhaps best described as ‘Malthus in, Malthus out’.

The mathematical physicist Sam Cole critically evaluated the structure of the World2 and World3 models, the main distinction between the models was that there were three times as many equations in World3 but that the quality of data underpinning these equations had not improved. In World2, Forrester had only used 2 data points, as Cole explained, ‘the only real data which appear to have been used in Forrester’s model are the world population figures for 1900 and 1970. For the most part the model is based on his vision of how the real world operates.’ In World3 there were numerous assumptions used such as inferring time series data (data measured over time) from cross-sections (data measured at one point in time across countries).

When it came to sensitivity analysis, Cole and the mathematician Robert Curnow found that the model was highly sensitive to the parameters chosen and that the model specifications were the ones that showed collapse: the assumptions drove the results. There was nothing in the model to prevent collapse, there were no adaptive feedback processes embedded. The model made pessimistic assumptions. If the model was run backwards then ‘Forrester’s guesswork’ led to unrealistic predictions about the past. If the model started in 1850 instead of 1900, then the model would show collapse in 1970. Clearly, the world had not collapsed in 1970, and this would invalidate the internal consistency of the model and its results. The model also led to inconsistent results if there was a split between developing and developed countries.

Subsequent analysis by Cole and Curnow, published in Nature, showed that the World2 and World3 models were highly sensitive if run backwards and led to some unrealistic conclusions about the past. The main implication was that the parameters of the model were unstable and that this was driven by the assumptions. These criticisms of the model are reflective of wider criticisms, such as other critiques published in the journal Nature, that also highlighted the sensitivity of the models to the assumptions and that it was ‘too early for policy conclusions’ based on the World2 and World3 models.

Economists on the Limits to Growth

LTG did not go unnoticed within the world of economics, especially those who had studied economic growth. Jay Forrester, an MIT professor in the Sloan School of Management, saw economists as the biggest critics of both his World Dynamics and the LTG. Given that the LTG was essentially an MIT affiliated study and also that the leading authorities on economic growth were either based at MIT (Robert Solow) or had studied there (William Nordhaus and Martin Weitzman), it effectively became an MIT internal dispute; or more accurately, an internal dispute of MIT’s Sloan School of Management. Robert Solow later recalled that: ‘I thought that its message was mostly hot air. I have not changed my mind about them’. Although he later felt that the situation may have changed given the resource intensive economic growth experienced by India and China.

The opening line of Solow’s initial views on the LTG were that his MIT colleague was ‘either the Christopher Columbus or the Dr. Strangelove of this business.’ Solow quoted an interview that Forrester had given where he had defined the composition of his ‘problem-solving group,’ they were to be made up of successful professors but it should purposefully exclude social scientists because they ‘always want to get to the bottom of a particular problem. What we want to look at are the problems caused by interactions.’ In this view, Forrester did not wish for accuracy on particular problems but wanted to get the interactions right.

Given this disdain for social sciences, it is then perhaps unsurprising that World Dynamics, effectively a book on economic growth, contains only 13 references and none of which are to any work by economists but instead were primarily self-citations of Forrester’s own work. It is somewhat curious though, given the renowned expertise in MIT at the time, that Forrester would not consult his colleagues in the economics department or even to use his work to chastise their efforts. Perhaps greater cross-campus dialogue could have helped inform the World2 and World3 models.

The main argument that Solow presented in his 1973 article were that the assumptions Forrester made drove his results, change the assumptions and the conclusions would change too. For example, Solow highlighted the assumption that population growth was positively correlated with income, something which went against the lived experiences of countries undergoing demographic transition. The model was seen as parameterised to show collapse.

Economists continued to engage with the various iterations of the LTG. One of the most vocal critics was William Nordhaus who highlighted how the underlying LTG model overlooked existing economic theory, invented concepts de nouveau, and set model parameters so that it only showed collapse. An initial critique by Nordhaus published in the Economic Journal in 1973 argued that changing some key assumptions in Forrester’s World Dynamics model changed the outcome dramatically; Nordhaus similarly dismissed updates to the LTG as ‘Lethal Model 2’.

Countering some of the claims was evidently still on Solow’s minds when he gave an address to the American Economic Association in December 1973, acknowledging that he, ‘like everyone else, [had] been suckered into reading the Limits to Growth’. This led to a series of studies that incorporated exhaustible (non-renewable) resources more explicitly in models of economic growth. It also led to the foundations of the economic approach to sustainable development and inclusive wealth. These focused on achieving intergenerational equity via the reinvestment of rents from exhaustible resources into reproducible capital (such as buildings, machine and tools).

Limits to Growth may have been partly right

In the various critiques of the LTG, the focus predominantly was on non-exhaustible resources as the LTG had placed greatest emphasis on this aspect. Nordhaus was sanguine about the pollution aspect of the LTG model stressing that growth need not result in increases in pollution (in terms of toxic gases) but placed less stress on global warming/climate change prognosis of the LTG, noting how ‘the cost of greenhouse warming in the middle of the next century (or of policies to slow greenhouse warming) would be in the range of 0 to 2 percent of world income.’ Similarly, Martin Weitzman focused exclusively on the exhaustible resource constraint set by the standard model of the LTG. Weitzman assessed the likelihood that the world was running out of resources and came to the optiistic conclusion that resource scarcity was not binding given that technological change was ‘maybe 40 times larger than the required adjustment for depletion of exhaustible resources.’

This narrative that economists had won the debate with LTG also overlooks a contemporary response to Nordhaus from Forrester and colleagues.² In their response, Forrester and colleagues argued it was Nordhaus who was wrong, and it was he who had misinterpreted the model. For Forrester, the systems model should be analysed as a whole and not broken into subcomponents, as was done by Nordhaus; because this was how the system interacts, and it is the interactions that drive the results. While Nordhaus had portrayed Forrester as a neo-Malthusian peddling a disproven theory; Forrester embraced this identity and instead accused Nordhaus of misunderstanding Malthus.

Figure 4 Atmospheric Concentrations of CO₂ (parts per million), 0-2023

_{Note: Horizontal dashed line at 280 ppm indicate the pre-industrial average atmospheric CO2 concentration. Horizontal dashed line at 350 ppm indicates the safe level of CO2 in the atmosphere (crossed in 1989).}

What these critiques and responses to the LTG overlooked, however, was one key aspect in the second scenario which emphasised pollution as a constraint on growth, primarily CO₂ emissions. The LTG predicated that if CO₂ emissions continued to grow at 0.2% per annum and if fossil fuel consumption continued unabated, then atmospheric CO₂ concentration would reach 380 parts per million (ppm) by 2000 ; remarkably the recorded figure in 2000 was not far off this, reaching 370 ppm; as shown in Figure 4.

This rise in CO₂ concentrations was not a foregone conclusion as the LTG noted, ‘if man’s energy needs are someday supplied by nuclear power instead of fossil fuels’. However, the LTG warned of global warming ‘by the laws of thermodynamics’ and that ‘thermal pollution may have serious climate effects, worldwide, when it reaches some appreciable fraction of the energy normally absorbed by the Earth from the sun.’

Degrowth and Limits to Growth

The LTGs enduring appeal comes from offering a coherent story that links environmental degradation, resource use, inequality and economic growth into a single framework. Even if the forecasts are disputed, the narrative remains powerful. In his 2025 book Slow Down or Die, Timothée Parrique structures his chapters around the ecological, social, and political ‘Limits to Growth’. Ultimately, he sees LTG as prophetic concluding that though it was ‘extremely controversial at the time, its ideas find almost universal agreement today.’ Other degrowth studies draws inspiration from LTG, or rather an ‘advanced reincarnation’ of LTG as Giorgos Kallis and Hug March see it. Recent degrowth literature reviews show that there is limited engagement with the concepts within mainstream economics. Perhaps the reason why there has been limited engagement within economics is because the LTG predictions were heavily criticised by economists and other scientists. From the perspective of economists, the predictions of the World3 model were not born out and thus are not worthy of debate.

The main issue appears to be different perspectives on the LTG debate. For economists, the debate was won and there is therefore no need for further engagement with proponents of LTG (degrowthers). For systems thinkers, they too have been vindicated. Some found that the LTG model fitted data 30 years on or even became more relevant over time, and there have been studies updating the LTG models. The most recent update to the World3 model shows a similar overshoot and collapse outcome to the original model. Effectively, this is an unsettled scientific controversy where recourse to facts or theory cannot resolve the controversy as both sides see the world in different ways.³

There are lessons from another recent scientific controversy, that over ‘zero-Covid’ during the 2020 Covid pandemic. The Covid-19 response was influenced by epidemiological modelling of the pandemic and the original models assumed no mitigation. Once such forecasts become public they can change behaviour, making the original forecast a moving target. Economists know this problem as the Lucas critique: predictions can alter the very system they seek to predict. There are direct parallels here to the LTG as a model was built, with various assumptions, and it revealed a systemic danger and this danger was not visible in conventional thinking, therefore policymakers must change course.

One of the most influential Covid-era studies argued that elimination rather than mitigation would produce the best outcomes for both health and the economy. The study became a touchstone for advocates of Zero-Covid, yet the policy was never sustainably achieved and later assessments of the evidence proved more mixed than many advocates initially claimed.⁴ A subsequent BMJ review found evidence of advocacy bias in parts of the Covid literature, with researchers often emphasising the benefits of preferred policies while paying less attention to trade-offs and uncertainty.

Both the Covid-19 and degrowth debates exhibit a common pattern in which model-based warnings become advocacy movements and once participants become advocates, evidence is often selected asymmetrically. For example, in The Guardian op-ed cited above the conclusion that we are entering ‘hothouse earth’ territory was based on a commentary that uses the SSP5-8.5 scenario which is no longer regarded by many researchers as the most plausible business-as-usual pathway. In fact, many climate researchers now regard SSP5-8.5 as ‘implausible based on trends in the costs of renewables, the emergence of climate policy and recent emission trends’. The issue is not that SSP5-8.5 was impossible, but that an extreme scenario gradually acquired the status of a baseline forecast and once a model becomes an advocacy tool, there is a natural tendency to focus on the most alarming trajectories rather than the most probable ones. Further the argument that growth has not alleviated poverty has also ignored the fact that economic growth in China and India has lifted more people out of poverty than conventional redistributionist policies as advocated by degrowth.

Fifty-four years after LTG, prominent economists and degrowth advocates once again tell us that growth is a doomed strategy. The remarkable thing is not that these arguments persist, but that they continue to rely on the same intellectual architecture: computer models, systemic risks, warnings of overshoot, and calls for urgent transformation. The question is not whether there are limits to growth but whether we can discuss those limits without allowing models to become articles of faith.

None of this is to say that growth is a perfect solution. The challenge is not to choose between growth and degrowth, but to think more carefully about what we mean by prosperity and how it should be measured. We need less advocacy bias and more scientific rigour in public discourse if we are to solve the genuine challenges that societies face today. I explore these questions further in my forthcoming book on Adam Smith and inclusive wealth.

1. The Science Policy Research Unit at the University of Sussex. The resulting report, Thinking about the Future, was published in the UK in 1973 and also appeared in the US with a more punchier title Models of Doom; the US edition also included a reply from the Meadows et al. team.

2. It is alleged that Forrester was not given a right of reply to the Nordhaus critique by the Economic Journal: M. Myrtveit (2005). The World Model Controversy. Working Papers in System Dynamics. Bergen Open Research Archive.

3. See, for example, H. T. Engelhardt Jr and A. L. Caplan (eds) (1987). Scientific Controversies: Case Studies in the Resolution and Closure of Disputes in Science and Technology. Cambridge: Cambridge University Press.

4. Revisiting the original analysis from the Lancet, shows that elimination countries performed slightly better in 2020 at -3.56 % compared to -4.65 % for non-elimination countries, but in 2021 elimination countries rebounded slower than non-elimination countries (3.41% v 6.70 %): see McLaughlin and Beck 2025). Managing and mitigating future public health risks: Planetary boundaries, global catastrophic risk, and inclusive wealth. Risk Analysis.