
Diane Coyle on the Soulful Science 04/14/2008
What this covers
Diane Coyle and Russ Roberts discuss the state of modern economics, examining where the discipline has genuinely advanced and where it has fallen short. Their conversation centers on growth—how it is measured, what drives it, and why policy makers often draw the wrong conclusions from economic data. Coyle argues that economics has made real progress through better datasets, sharper formal reasoning, and applied microeconomics in domains like competition policy and market design. Yet the field's largest failures—particularly in development policy and macroeconomic growth regressions—stem from applying tools beyond what the underlying data can reliably support. She contends that institutions, human capital, and technology, not capital accumulation alone, explain growth differences between nations.
The conversation moves across several domains beyond growth theory. They examine how behavioral economics has enriched the discipline by incorporating psychology, how technology and globalization produce superstar effects that drive inequality, and why happiness research yields less reliable insights than cross-sectional studies of life satisfaction. Coyle discusses why development aid often misses transformative private-sector innovations—offering mobile phones in Africa as an example of untracked but profound change. They also address the reliability problem directly: macroeconomic regressions claiming causal links between specific policies and growth rates are unreliable when applied to countries with poor data collection. The discussion includes contentious terrain on inequality measurement, redistributive taxation, and whether global inequality figures actually inform policy usefully. Coyle also reflects on how formalism serves economics—as a tool for clarity and computation—and how the field's culture, particularly its aggressive intellectual style, has shaped who enters and stays in the discipline.
Coyle argues that modern economics has genuinely advanced—especially through better data, formalism that sharpens thinking, and applied microeconomics like competition policy and market design—but that its biggest failures (development policy, crude growth regressions) stem from misapplying tools beyond the limits of the data, and that institutions, human capital, and technology, not capital accumulation alone, drive growth.
- Better datasets (e.g. Maddison) revealed that growth comes from technology and human capital, not just labor and capital inputs
- Formalism is valuable for clarity of thought and econometrics, but macro growth regressions are unreliable given poor data quality
- Applied microeconomics (competition policy, market design, auctions) is where economics has delivered real public-policy payoffs
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
GDP cannot accurately capture qualitative improvements in living standards over long periods, because the very concept of output changes over time and new goods appear—illustrated by a woman in a TV reconstruction of 1900 life who was driven to despair because shampoo hadn't been invented until the 1930s.
“the concept of output itself changes over time... it can't possibly take account of improvements in living standards that a qualitatively different”
Markets are themselves one form of institution and work well or badly depending on the whole array of surrounding institutions; where property rights are well enforced and there is a good legal system markets work better, which is why markets automatically don't work well in many developing countries that lack those institutions.
“markets are actually one form of institution and so they may work well and they and they may work badly depending on the whole array of institutions in the economy”
The spread of mobile phones in African countries—driven by profitable private companies outside the development community's radar—has been transformative, turning an unemployed laborer into a self-employed artisan by letting them advertise services and find work, edging people from the informal into the formal economy, spawning clusters of micro-businesses (SIM card sellers, drinks vendors), and now enabling profitable mobile banking that brings financial services to people with nowhere safe to keep money.
“having a mobile phone... somebody from an unemployed laborer in a Township to a self-employed artisan because they can advertise their their services find work it's edging people into the formal economy from the informal economy”
The core difference between environmentalists and economists is that environmentalists extrapolate present conditions into the future, while economists assume that shortages cause prices to adjust and behavior to change—illustrated by Paul Ehrlich losing his bet with Julian Simon when metal prices fell as technologies and behavior adjusted; the bet works over roughly ten-year horizons but not always short-term.
“environmentalists extrapolate from space conditions into the future an economists tend to assume that if there's a shortage or if there's a problem something will adjust prices and markets will reflect what's happening and behavior will adjust”
Correct development conclusions change over time because context changes—the protectionist development of Japan and Korea in the 1950s-60s is not a valid template today because the world economy and the goods being traded were far simpler then, so a developing country today cannot replicate that path into the more complex modern goods market.
“the crowd economy in the 1950s and 60s with a very different place than it is there and the goods that people were making a buying rule much simpler”
Economics has a self-reinforcing male outlook and gladiatorial culture: the proportion of women declines at every step beyond high school, partly because the field's aggressive, formalism-heavy, intellectual-pride culture (e.g. the Chicago workshop system) easily discourages those who don't 'get it' straight away, and because young women see economics publicly represented by people talking about financial markets going up and down—a turn-off.
“why so few women the economics and the numbers the proportion declines the third you grow you go through the professional training”
The most useful thing economists can do is make policymakers pay attention to the unintended consequences of interventions—captured in Thomas Sowell's 'and then what?' and Bastiat's distinction between the seen and the unseen—because nearly every government intervention produces unintended consequences.
“one of my hopes for economists in the future is that we can start to make policy makers pay more attention to what the unintended consequences might be I think it's probably our best thing”
GDP and broader indices like the Human Development Index (life expectancy, infant mortality, clean water, literacy) are highly correlated up to about $20,000 per capita, so for most of the world's countries GDP growth remains a good policy target because it brings along the other things people care about.
“they're the highly correlated with GDP at least up to a level of about $20,000 per capita so for a lot of countries in the world growth economic growth GDP growth is still a really good policy target”
Modern growth theory focuses on two factors that intuitively explain why economies grow—technology that makes more efficient use of inputs, and people with skills and education—and when human capital is added to models you get feedback loops that produce snowballing endogenous growth, which explains rare but dramatic 'growth miracles' like the East Asian countries in the 70s-80s or Ireland.
“modern growth theories focus on the two things that intuitively make complete sense about why economies grow what is the technology... and the others the people and their level of skills and education”
Applied microeconomics has delivered real payoffs: better competition/antitrust policy informed by theory and evidence has produced more competitive markets with measurable gains in productivity and consumer prices; and market design—replacing 'beauty contest' license allocation (e.g. for spectrum or oil exploration) with auctions—generates government revenue, excludes petty corruption, and directs licenses to whoever will use them most efficiently.
“the gathering of evidence in those cases has improved so much that actually we have more competitive markets than we used to with a payoff in terms of productivity and the prices that consumers pay”
As world population has grown, world poverty has declined not just in relative terms but in absolute terms in many areas—an extraordinary and underappreciated moment in human history that runs against the prevailing pessimism, driven heavily by China and India.
“as world population has grown world poverty has not just declined in relative terms but in many areas in absolute terms which is really extraordinary”
Superstar effects explain rising inequality in many professions: technology and globalization let people who are only slightly better than rivals leverage their skills into a worldwide market (e.g. a movie star whose films show everywhere, or top lawyers and doctors), capturing a disproportionate share of available revenue, with an added effect of people paying a premium specifically to see the very best.
“some people who because they can leverage their skills into a wide market can make extremely high incomes and the classic example is a movie star”
Economics has become much less about rational choice than it was decades ago, having been enriched by importing insights from psychology—particularly behavioral finance, which gives a richer understanding of how markets and asset markets behave and how people make financial decisions affecting their distant future, given our psychological difficulty with such calculations and our risk and loss aversion.
“economics is much less about rational choice than it was when when you and I were in graduate school and it's been a very fruitful I think the insights from psychology applied to financial markets”
Reported happiness has not risen much over time despite large measurable improvements in living standards, but this evidence is weak because happiness data and GDP are different animals: GDP can grow without bound while happiness is a bounded emotion measured by surveys with only a few choices, so the regressions don't tell you much; cross-sectional studies more reliably show happiness depends on having a job, being married, and being part of a community.
“GDP per capita is the time series that can grow without bound over time happiness is an emotion we doesn't behave in the same way”
Redistributive tax policy can have a self-defeating market effect: raising taxes on the wealthy and lowering them on the poor likely increases before-tax incomes for the wealthy because of competition to attract skilled people, thereby increasing measured pre-tax inequality and offsetting some of the equalizing effort.
“if you raise taxes on the wealthy and you lower them on the poor you're going to like you're likely going to increase before tax incomes for the wealthy because of market effects competition to get these the skilled folks”
Macroeconomic regressions of growth are unreliable because they use broad-brush, inaccurate measures—especially for developing countries, where a state that can't deliver water or energy to its people probably can't collect good statistics either—so regressions claiming e.g. 'build more roads and growth goes up X percent' are not helpful.
“if you think about really how statistics are collected in a developing country that can't deliver energy or water to its people it's probably not going to be very good at collecting statistics either so for all these reasons I'm quite skeptical about using regressions mechanically”
Underlying inequality swings driven by technology and globalization can take a long time (a generation) to reverse, as happened in the late 19th century when public education and the spread of high schools eventually broadened human capital and narrowed inequality; policy should focus on ensuring nobody is immiserated and that people's real incomes don't decline while society gets richer.
“the skills that were in short supply were filled by the public education system and development of high schools throughout the early 20th century and when a wide-range the population had better human capital inequality narrows”
Global inequality figures are misleading and foolish to act on because they essentially compare extremes (e.g. Bill Gates to the poorest person in Benin) while ignoring the distribution in the middle; what matters is individual identities, illustrated by the fact that the richest Brazilians are much better off than the poorest French people whose taxes fund overseas aid to Brazil.
“if you look at national averages Brazil looks like a much poorer country in France and France keys aid to Brazil but actually the richest Brazilians are much better off than the poorest French people whose tax money is paying for the overseas aid”
More people tends to lead to a better standard of living, following Adam Smith's idea that the division of labor is limited by the extent of the market and Paul Romer's insight about the power of size; there are no examples of economies with shrinking populations where real income per capita has grown.
“I can't think of any examples of economies with shrinking populations where real incomes per capita has grown”
The economics of different countries (continental Europe vs the Anglo-Saxon world) has converged dramatically, partly due to subject changes and partly due to the English language—economists in Italy, Germany, and France now publish in English and join the same intellectual current—so there is no longer the kind of chasm between schools that existed in the 1930s.
“the kind of economics that economists in Italy or Germany or France is now much more much closer to the kind of economics the people in the UK or us would”
Cross-sectional snapshots of inequality are misleading because they compare different people at different points rather than tracking individuals; for example, falling/flat wages for high-school-only graduates in the US don't tell you what happened to people 25 years ago because the share of people with only a high-school education has shrunk dramatically.
“those can be very misleading they're not the same people and it looks as if certain people are being made poor”
The UK's competition authority works better than US analogues because it is an independent, technocratic body embedded in law: members serve eight-year non-renewable terms, are appointed through a civil-service recruitment process rather than partisan balancing, cannot have active political affiliations, and adjudicate mergers as economic decisions—free from the partisan Republican/Democrat split that governs US commissions.
“the fact that it has become seen to be an economic decision as a technocratic decision if you like that's now embedded in law essentially led on this side of Atlantic to a much tougher competition policy”
US antitrust enforcement is flawed and politicized: large-scale cases launched roughly every decade against a leading firm (IBM, Microsoft, the cereal industry) tend to last so long that the target is out of business or irrelevant by the time the suit ends, and antitrust is often used by losing competitors lobbying politicians to crack down on successful rivals who should be left alone.
“we do these large-scale every ten years we decide to take a large firm and beat it up and by the time the suits over there usually out of business or unimportant anyway and we did this with IBM we tried to do with Microsoft”
Formalism (mathematics) is valuable in economics for two reasons: it forces logical, imperative thought and clarity, which good economists should then be able to translate back into plain language; and it enables the econometric and computational techniques behind genuinely useful applied policy like spectrum auctions, competition policy, and modeling transportation systems.
“one important reason is the clarity of thought that it brings... because it forces logic imperative thought on your understanding of the world”
Development policy is an area where economists have not covered themselves in glory in the post-war period; politically motivated fads using crude regressions justified pouring in money that was spent badly and became a curse for many poor economies—though development economics has at last moved to a more sophisticated debate with greater realism about the limits of external capacity to lift economies out of poverty traps.
“development policy is an area where economists are certainly not covered themselves in glory in the post-war period and as you say there have been all kinds of fads they have been very politically motivated”
The Solow model's focus on K (capital) and L (labor) led to a mindless policy conclusion that poor countries could be made rich simply by adding investment and education; the theory may have been roughly right but implementation was inherently flawed because of incentives and poor governance.
“one of the tragedies of the solo model was that by focusing on K and L it led us to this very mindless... idea that we need more capital better labor... and we just need to to add investment and education and we'll make poor countries rich”
There is evidence from the evolutionary literature that humans have a sense of fairness, so societies that become too unequal can suffer instability; but while extreme inequality (e.g. China's regional divergence) may be clearly unsustainable, there is no clear evidence of where 'too unequal' lies among Western democracies, and inequality talk in the US functions mainly as a rhetorical device rather than driving aggressive policy action.
“there is evidence that societies that become too unequal not not stable societies and suffer a group Enel t because of it I'm not sure there's clear evidence of what too unequal is”
Environmental 'genuine progress'-style indices are constructed to subtract social bads (pollution, crime) from GDP without ever adding the positives of growth (new products like home computers, price declines, increased variety), so they are designed to never show improvement and put people in a pessimistic frame of mind; it makes more sense to publish an array of separately-weighted indicators (as Australia does after public consultation) than a single index.
“they have constructed them so that you can't get any improvements so they have got their countrymen a very pessimistic frame of mind”
Grand development schemes usually fail, so the better approach (following Easterly's insights) is small-scale, incremental steps rather than big top-down interventions.
“I'm a big fan of of easterlies insights and and the belief that grand things usually fail so I'm all for small-scale small steps we've tried to big things”
Despite heroic UK government attempts over the past decade to equalize incomes through tax policy and tax credits (modeled on the American EITC) with very large redistributive effects, incomes still became more unequal because underlying income inequality increased—driven by technology and superstar effects in many markets.
“the redistributive effects of taxation and Welfare has been extremely large but even so incomes has become more unequal because the underlying growth and income inequality has increased”
Creative destruction is tolerated well in the United States largely because recovery from it is so fast; if recovery were slow, people would likely be far less enthusiastic about it.
“one of the reasons creative destruction is tolerated very well in the United States by most people and advocated by many is because the recovery is so fast”
Most people cannot meaningfully assess their relative income position: they know their salary within ~10% but struggle to value their benefits, have little idea what income percentile they occupy, and cannot know how their relative position has changed over time—so claims about relative inequality heard in the political arena are largely unverifiable by individuals.
“my a puzzle always is I always ask people if they know their income most people know their annual salary within say ten percent they struggle to tell you what their benefits are worth”
An extended-warranty market investigation found it was not competitive because warranties were sold opportunistically at the point of purchase ('would you like this for just an extra fifty pounds') at poor value; the authority's remedy of displaying warranty prices alongside item prices was too weak, and in hindsight they should have banned selling warranties at the same time as the insured item to give buyers space to reflect.
“they were not very good value so we proposed remedies like that which were eventually slowly implemented with hindsight I think we should have banned them from being sold at the same time as the item that they were insuring”
In network and technology industries the best form of competition may be the next new technology coming along, complicating standard competition analysis.
“in technology industries do you think the best form of competition is another new technology coming along”
Economics genuinely improved monetary and fiscal policy: in the 1970s monetary policy was an ideological decision, but there is now much greater consensus on how to run monetary and budget policy, producing better outcomes over the last 25 years.
“in the 1970s it was an ideological decision what he'll make food coming policy would be and we now have much greater consensus about running have a monetary policy and budget policy”
Growth is the central question in economics and the one Adam Smith started with.
“growth is obviously the central question in economics and it's the one that Adam Smith started off with so long ago”
Working in the competition authority on nine cases across the economy was a fascinating education in how businesses and markets really work, because every market is different.
“it's been a fascinating education in how businesses markets really do work... every market is different”