What this covers

Diane Coyle joins Russ Roberts to argue that economics must abandon its fixation on isolated rational actors and single measures of value if it hopes to remain relevant to real-world policy. The conversation moves through concrete failures: supply chain opacities that firms themselves cannot map, machine learning systems that embed utility function choices without addressing social choice theory, and the 2008 financial crisis, which forecasters saw coming but chose not to heed—a case of motivated inattention rather than absent signals. Coyle's core complaint is that mainstream economics treats humans as atomized "cogs," ignoring Complexity / Interconnection, network effects, and the role of social norms, especially in the digital economy. It also claims scientific objectivity while embedding value judgments—a pose that feeds expert backlash.

The substance turns on two interconnected problems. First, the Money Metric / Single Scalar of Value flattens what cannot be flattened: people care about whether Stonehenge stays in place even if they never visit, and they want their society to preserve things they individually consume no more of—a gap between Citizen vs Consumer Preferences that GDP captures not at all. Second, Methodological Individualism misses how place, rootedness, and collective goods (public transport, libraries, courts) shape what people value. The conversation notes that welfare economics has not been seriously rebooted since the 1970s, yet structural economic change—especially around intangibles and digital transformation—demands it. A persistent tension emerges: decisions still require frameworks, yet the impulse to collapse everything to one scale may itself be the problem. Coyle gestures toward Amartya Sen's Capabilities Approach as a corrective—expanding the range of what counts as human flourishing—without claiming it solves the Incommensurability of Values.

Sharpest takeaway

Coyle argues that mainstream economics over-relies on methodological individualism, objectivity claims, and a single money-metric of value, and that to stay relevant it must incorporate social interaction, value pluralism, and a rebooted welfare economics suited to the digital economy.

  • Economists treat people as atomized 'cogs' (rational maximizers) and neglect network effects, social norms, and collectively-provided goods.
  • Economics poses as objective but embeds value judgments, contributing to a backlash against experts.
  • Reducing all value to one money metric fails for cultural, civic, and intangible goods, so the discipline needs more diverse methods.

The argument · threads7 threads · 37 claims
0.76

Economics falsely claims objectivity and impartiality while embedding value judgments in its own analysis.

5 pointscentrality 5/5
  • Economics has adopted a pose of objectivity and impartiality—comparing itself to dentistry or plumbing—but its own modes of analysis embed value judgments, and the backlash against experts is partly a reaction to the false claim of top-down impartial objectivity.

    “part of the backlash against experts is a backlash against the claim to be impartial and objective and deciding the best thing for everybody in this top-down way, when it isn't the case.”

  • Post-financial-crisis unconventional monetary policy was claimed as an objective, technocratic decision by independent central banks for the best economic outcome, but it had major distributional consequences that central banks were slow to acknowledge and that laid the groundwork for large shifts in the distribution of wealth and income.

    “it was very much claimed to be a decision taken by independent central banks to deliver objectively the best outcome... but it had great distributional consequences, and it took a long time... for the central banking community... to 'fess up to that. And, it has laid the groundwork for massive change in the distribution of wealth and income”

  • Replacing GDP by directly measuring people's wellbeing is dangerous because well-meaning economists can end up imposing wellbeing on people on a simplistic scale—'we will do wellbeing to people because we know this is what's good for you'—which is just another urge to reduce everything to one dimension and one set of criteria.

    “I'm a little bit concerned about saying, 'We'll just ignore growth, and we'll go directly to measuring people's wellbeing'... you can end up with very well-meaning economists saying, 'We will do wellbeing to people because we know this is what's good for you'... Equally, that's an urge to reduce everything to one dimension”

  • Economists are trained as utilitarians, using a single money metric to pile everything into one number for go/no-go decisions, but this reductionism feels alien to many people who hold that aesthetic, spiritual, and identity values are genuinely multidimensional and cannot be collapsed into one scale—creating a core dilemma since policy still requires a decision-making framework.

    “our presumption that you can find a metric, and we use money--the money metric--and pile everything into that one number to get a no-go decision... seems alien to many people... they're very strong resistance to the idea that the aesthetic value, and the spiritual value, and the identity value cannot be considered separately.”

  • Machine learning systems increasingly used in the private and public sectors to make consequential decisions about people's lives are literally utility maximizers—set up to maximize objective functions or minimize loss functions—chosen by computer scientists, despite all the unresolved social choice problems economists know.

    “a lot of machine learning systems increasingly being used in the private and the public sector to make decisions that affect people's lives significantly are actually utility maximizers. They're set up to maximize objective functions or minimize loss functions.”

0.72

Societal value and well-being are multidimensional and cannot be reduced to revealed individual preferences or a single monetary metric.

4 pointscentrality 5/5
  • People are well aware that their day-to-day personal preferences differ from what they want for their society: they may choose to watch soap operas while still wanting religious and children's programming to exist, and want Stonehenge preserved in place even if they will never visit—so societal value is not reducible to revealed individual consumption.

    “people will choose to watch soap operas or movies that get shown on TV, but they want there to be religious programming and children's programming, even if they're never going to watch it themselves. They're good things to have.”

  • Universal Basic Income is the ultimate individualist policy, but individual cash transfers cannot buy the collectively-provided goods—transport networks, public schools, healthcare—that matter most for communities and the least well off.

    “what you can't do with an Universal Basic Income is buy collective goods. And, to the extent you care about communities and improving the chances of those who are the least well off, then it's often those collectively-provided goods that matter a good deal”

  • Although economics has been too reductionist, growth in advanced economies over the past several decades has not meant consuming more material 'stuff'—the change in GDP has been almost entirely intangible, with economic value increasingly assigned to intangibles rather than physical goods.

    “if you look at the increase in the consumption of material in the advanced economies in the past 20, 30, 40 years, we haven't been consuming more stuff. We've been assigning economic value to intangible stuff... the change in GDP was all intangible. It wasn't about stuff.”

  • Moving Stonehenge to a transport-connected exhibition center would score better on conventional economic value and even on visitor access, yet the lesson for students is that economics isn't always the answer—people value things being in their authentic place even if they will never visit.

    “propose to them that we move Stonehenge from the middle of the countryside to the exhibition center in Birmingham... that are obviously going to have much better economic value... the lesson I want them to draw from that, of course, is that economics isn't always the answer.”

0.57

Government functions as insurance against unforeseeable adverse risks, requiring sudden large-scale interventions that reshape what policy deems possible.

7 pointscentrality 3/5
  • The aftermath of Brexit is showing in a collapse in UK export volumes—which recovered less than other OECD countries—and rising transaction costs of trade that especially burden small businesses, making it look like the earlier prediction of a roughly 4% loss of GDP will prove fairly accurate, though COVID and the Ukraine conflict make this hard to disentangle.

    “The aftermath is manifesting itself in the collapse in export volumes, and it looks like those earlier predictions of a 4% loss of GDP will prove pretty accurate.”

  • Before the 2008 crisis, clear strain signals (house-price-to-earnings ratios rocketing, liquidity strains, volatility indices) were present for about a year, but forecasters defaulted to the comfortable 'slow resolution' scenario over the crisis scenario, so the character of attention paid—not absence of signals—explains why nothing was done to avert it.

    “either all of these strains will resolve themselves over time gradually... or there'll be a crisis. And, we all then defaulted to talking about the slow resolution rather than the crisis.”

  • Using the old cobweb model as a lens, current shortages may reflect a divergent rather than convergent dynamic because physical limitations prevent quantities from adjusting quickly, producing large price changes without a supply response; understanding where the bottlenecks are is therefore an urgent issue.

    “if you thought about it in terms of something like the old cobweb model for agricultural output, I worry that we might be in a divergent context rather than a convergent context, because the physical limitations mean that those quantity adjustments aren't happening.”

  • A Peterson Institute study of vaccine-manufacturing supply chains shows a more optimistic case: the complex supply chains needed to manufacture new kinds of vaccines at scale were created incredibly quickly, due to personal relationships and urgent new government-set structures, demonstrating shortages can be fixed if we know where they will be.

    “the adjustment to the need to create complex supply chains--to manufacture at scale the new kinds of vaccines--that happened incredibly quickly. And, for all kinds of personal relationship reasons and the urgency of the new structures that were set up by governments, that happened.”

  • A key pandemic lesson is that governments can undertake sudden, large-scale interventions—something previously outside the universe of the possible in much policy thinking—even though some of those interventions worked well and some were mistakes.

    “governments can do things--that kind of sudden and large intervention that governments made... simply the fact that that could be done. And, that had been, I think, outside the universe of the possible in much policy thinking beforehand.”

  • Beyond the extraordinary achievement of vaccines, governments have not grasped what is needed to prepare for the next pandemic—covering the range from public health systems and warning systems to underappreciated factors like clean air, which affects human capital and pandemic preparedness.

    “I don't think they've got their heads around it at all... clean air is quite important, and it isn't just about being warm and cuddly and hugging trees. Actually, clean air affects human capital and pandemic preparedness.”

  • A useful way to think of government is as providing insurance against adverse turns of fortune we could not possibly have averted ourselves; in the early 20th century that meant mitigating business cycles and an adequate welfare state, and today, with different kinds of risks, it may be time to think about government in a different social-insurance role.

    “what we really want governments to do is to provide some insurance against adverse turns of fortune that we couldn't possibly have averted ourselves... now, the risks that we face may be different kinds of risks. And so, maybe it's time to think about government in a different kind of social insurance role.”

0.54

Economic criteria are not always decisive in policy because people hold aesthetic, spiritual, and identity values that diverge from efficiency maximization.

3 pointscentrality 4/5
  • A major divide emerging across all Western economies is between highly educated, mobile people who move to big cities where high-skill jobs congregate and the majority who stay rooted where they are; this implies sometimes overriding efficiency criteria to ensure everyone has good public services, as part of living in a democracy.

    “the divide that's emerging is between highly educated people who are more mobile and go and live in the big cities where all the high-skill jobs are now congregating, and the majority of people who stay rooted where they are. So, this is not just a UK/Brexit problem. This is a kind of Western economies problem”

  • Brexit was a bigger shock to UK economists than the financial crisis because, unlike the crisis (which most economists had not studied), the profession thought a great deal about Brexit and reached a strong consensus that economic damage would result, yet voters said they did not care—demonstrating that economic criteria are not always the most important criteria in people's decisions.

    “the profession thought a lot about Brexit, and there was a very strong consensus that there would be economic damage caused. And, voters said, 'We don't care.' And so, it is a great lesson... that economic criteria are not always the most important criteria when people make decisions.”

  • The Brexit vote was stronger in 'left-behind' post-industrial areas where living standards had stagnated since the financial crisis and where the footprint of public services had shrunk—closed libraries, courts, centralized hospitals, cut bus services—so people experienced their towns as deteriorating shrinking places and blamed the EU even though it had not caused those declines.

    “if you're living in one of these towns or outskirts of the industrial cities, you'd seen the fabric of the place you lived deteriorate. The local libraries closed... So, they felt like shrinking places... it was what got the blame, and that was what drove the successful Brexit campaign.”

0.54

Economists lack adequate data and conceptual tools to understand complex economic networks, supply chains, and digital economy structures.

3 pointscentrality 4/5
  • Whether on the production side (supply chains) or the finance side, we lack the data and a good way of understanding the deep network 'wiring' of the economy, so we cannot tell whether risks are being diversified or concentrated or what these connections imply for aggregate outcomes.

    “That deep wiring, whether it's on the production side or the finance side, I don't think we have the data or a good way of understanding what that implies for collective outcomes, for aggregate outcomes in the economy.”

  • Supply chains are so opaque that firms themselves don't know their critical interdependencies—illustrated by fertilizer factories that cut production when gas prices rose, not realizing their carbon dioxide byproduct was a crucial input to the food packaging and preparation chain; if the companies don't know it, economists certainly won't.

    “the byproduct of that is carbon dioxide, which they sold on into the food chain. They didn't know they did that. They had no idea that their byproduct was a crucial input into food packaging and food preparation.”

  • The characteristics of the digital economy are changing in ways economists have not thoroughly explored, leaving 'monsters'—scary, unmapped features—that require updating economic analysis to understand.

    “the characteristics of the economy, digital economy, are changing in ways that we haven't yet thoroughly explored, and there are some scary monsters out there that we need to tackle and update our economic analysis to be able to understand better.”

0.50

Welfare economics needs fundamental rethinking because defining progress is genuinely difficult during periods of dramatic structural transformation.

2 pointscentrality 4/5
  • The core question of what it means for a policy to make things 'better' lacks good tools in economics; welfare economics has not been seriously revisited since the 1970s texts of Little and Sen, and given dramatic structural changes it needs a reboot.

    “the texts that people still refer to date from the 1970s: [I.M.D.] Little, [Amartya] Sen... So, it seems to me it needs a kind of reboot.”

  • The question of what counts as progress becomes salient mainly when things are changing dramatically; in stable decades, standard metrics like GDP growth and incremental life-expectancy gains are good enough, but during general-purpose technological transformation, pandemics, and geopolitical upheaval—which restructure jobs and dramatically change income distribution—defining progress becomes genuinely hard.

    “It becomes a pressing question at times like now, when the technologies are bringing about quite dramatic changes. They're general purpose technologies, so that's restructuring people's jobs and consumption opportunities; it's changing the distribution of income quite dramatically.”

0.40

Macroeconomic forecasting is inherently limited; treating forecasts as scenarios rather than predictions and integrating cross-disciplinary approaches yields better understanding.

3 pointscentrality 3/5
  • It is unreasonable to expect economics to forecast crises; macroeconomic forecasts are stylized tools for thinking about what might happen rather than actual predictions, and we might do better to call them scenarios rather than forecasts.

    “I wish people were more aware that forecasts of that kind--macro forecasts of that kind--are quite stylized tools for thinking about what might happen rather than actual predictions. We might do better if we even called them scenarios rather than forecasts.”

  • Early in the pandemic economists and epidemiologists produced separate forecasts and only later combined forces; despite each approach's limitations, integrating the two was necessary because of clear interactions between them, a lesson now embedded in good cross-disciplinary work.

    “the economists did their forecasts, and the epidemiologists did their forecasts. It took a while for them to combine forces. And, for all the limitations of each approach to forecasting--actually, you needed that combination”

  • Methodological individualism is not wrong in itself, but economics has over-emphasized it at the expense of insights about how people interact and influence each other—insights that are increasingly important given the pervasiveness of network effects and the power of social norms in the digital economy.

    “it's not that there's anything wrong with methodological individualism, because of course we do act as individuals, but I think there's been an over-emphasis on that at the expense of some of these insights from social interactions.”