What this covers

Ed Leamer joins Russ Roberts to argue that macroeconomics is fundamentally a discipline of pattern-seeking and storytelling rather than scientific hypothesis-testing. Leamer contends that standard econometric analysis rarely settles disagreement between economists—empirical precision does not drive actual belief—and that the field would be more honest acknowledging its dependence on compelling narratives about data rather than pretending to produce true or false conclusions. The conversation turns on a specific empirical claim: that housing and consumer durables, not Wall Street or business investment, drive the timing of recessions, appearing to weaken in eight of ten postwar downturns three or four quarters before the official recession begins. This pattern implies current policy is misdirected, since stimulus packages are designed around business investment while the actual mechanism lies in household spending.

Leamer walks through the 2008 crisis as a case study. He traces subprime lending through a Minsky pattern—from sound borrowers to speculation to Ponzi-stage loans—and argues the Federal Reserve's low real rates in 2002–2004 were the "teaser rates" that fueled it. Once rates rose sharply, housing volume collapsed and price appreciation stopped, ending the refinancing machine. He distinguishes this housing-sector credit crunch from the broader corporate credit stress of 2008, the latter being partly a return to normal pricing rather than evidence of broken capital markets. More broadly, Leamer observes that U.S. long-term growth has held at roughly three percent for 35 years despite wildly different policies, suggesting macroeconomic interventions lack empirical force on growth—a conclusion that undercuts both sides of the stimulus debate, which he frames as ideological rather than evidence-based. He also critiques econometric training in graduate economics for teaching technique without developing genuine economic thinking.

Sharpest takeaway

Leamer argues that macroeconomics is not a science but a discipline of pattern-seeking and storytelling, and that the housing/consumer cycle—not Wall Street or business investment—drives recessions, implying current policy is misdirected.

  • US long-term growth has been remarkably constant (~3%) for 35 years regardless of policy, so policy debates about growth lack empirical support
  • Housing and consumer durables weaken before recessions in 8 of 10 cases, making it a 'consumer cycle' not a 'business cycle'
  • The 2008 crisis was a Main Street (housing/consumer) problem amplified by policymakers' fear-inducing response, not a Wall Street credit-crunch causing the real downturn

The claims · ranked37 claims · weighted by value

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.

0.83

No sophisticated econometric (as opposed to simple fact-based) study has ever produced a new consensus in economics that wasn't there before simply because the empirical work was so decisive; equivalently, the standard errors that econometrics labors to produce don't drive anything economists actually believe—occasionally the sign of a coefficient matters, but that is rarely driven by empirical results.

factualhigh valuecontestednovelty 4/4durability 4/4· Ed Leamer

“whether there is any econometrics study I don't mean fact-based I mean econometrics sophisticated statistical analysis that has brought about a consensus in the profession that wasn't there before”

0.81

The subprime cycle followed a Minsky pattern: early loans go to borrowers who can pay principal and interest; later loans go to borrowers who can only service debt and rely on appreciation; finally, in the Ponzi period, loans go to borrowers whose income cannot even cover debt service and depend entirely on home price appreciation—exemplified by the '2/28' product (two-year teaser rate, 28 years of reality) that worked as a refinancing fee machine only while prices appreciated and collapsed once appreciation stopped.

causalhigh valueestablishednovelty 3/4durability 3/4· Ed Leamer

“the historian housing is one of these Minsky cycles in which early in expansions the loans are given to individuals who have the income go to service the debt but also to pay off the principal”

0.81

Economists should recognize the limits of their knowledge and admit that macroeconomics deals with complex, non-experimental human systems—so it is more accurate to describe the discipline as seeking patterns and telling stories than as testing evidence, and the standard for quality should be persuasiveness rather than scientific truth.

normativehigh valuecontestednovelty 3/4durability 4/4· Ed Leamer

“it'd be much more accurate to say that we are seeking patterns and telling stories rather than tearing evidence in other words it's a kind of literature that we're carrying on and the question isn't whether it's true or not the question is is it persuasive”

0.81

Graduate economics education teaches the 'grammar and syntax' of the discipline—how to do complicated econometrics and write papers that look scientific—but does not teach students how to think like economists or attend to the content of the message, which is a great indictment of the profession.

normativehigh valuecontestednovelty 3/4durability 4/4· Ed Leamer

“we teach people how to do the mathematics how to carry out a complicated econometrics how to write papers that look like they they are scientific but we don't teach people how to think like economist”

0.78

When econometric conclusions are highly sensitive to the assumptions used, a sensitivity analysis showing that sensitivity should place us in the 'we don't know' third logical category—the result is too sensitive to be believed.

normativehigh valuecontestednovelty 3/4durability 4/4· Ed Leamer

“when you do this econometrics you ought to do a sensitivity analysis to show how much it your conclusions depend upon the assumptions are you using and therefore if they're very sensitive then we ought to be in that third a logical position”

0.78

Macroeconomics is fundamentally an interventionist discipline, which is precisely why economists need to draw causal conclusions—but the available data are not up to that task, providing only temporal orderings (houses move first, then cars, then recession) to which a compelling causal story must be added.

factualhigh valuecontestednovelty 3/4durability 4/4· Ed Leamer

“we economists are are fundamentally an interventionist discipline... so we really are need to make causal conclusions unfortunately we don't have the data that are up to the task all we have these temporal orderings”

0.73

The fiscal-stimulus debate is not really about science—there is little empirical support for either side—but is fundamentally about whether America should be more like Europe (more government) or less, with each camp's position on the stimulus reflecting that prior preference rather than evidence.

normativehigh valuecontestednovelty 3/4durability 3/4· Russ Roberts

“the people who think we ought to spend two trillion extra one in America that's American role for a government role in America that's more like Europe's and then people who want us to spend less are saying I want to be less like Europe and that's what they're arguing about not not the so-called science”

0.73

The current global crisis is due to the overstatement of the precision with which economists and financial analysts claim to understand how the real world works; rating agencies in particular used a two-valued logic (true/false, safe/unsafe) when a third category—'we don't know'—was needed, and mortgage-backed securities should have been rated 'AAA-hypothetical' because the models were untested under stress.

causalhigh valuecontestednovelty 3/4durability 3/4· Ed Leamer

“the crisis that were in now on a global scale is due to the overstatement of the precision with which economists and financial analysts understand how the real world works”

0.73

What we experience is a consumer cycle, not a business cycle: in 8 of 10 postwar recessions the first expenditure component to weaken is spending on homes, then consumer durables (especially automobiles), and only after the recession officially begins does business cut equipment/software (short-lived assets) and finally long-lived structures—and the recovery follows the identical order, implying policy should target consumer spending rather than business investment.

causalhigh valuecontestednovelty 3/4durability 3/4· Ed Leamer

“it's it's not a business cycle that we experience it's a consumer cycle”

0.73

The reliable recession precursor is housing volume (sales and building), not price: historically home prices adjust very slowly because owner-occupants are reluctant motivated sellers (e.g., Southern California 1991 saw volume drop ~50% within a year but prices decline only 4-5% annually over five years), whereas this time was different because foreclosures came early and banks became motivated sellers, producing price declines even before the recession began.

causalhigh valuecontestednovelty 3/4durability 3/4· Ed Leamer

“what I'm emphasizing as a predictor recession isn't that home prices because of the way that home prices are determined historically which is sluggish and not a very reliable indicator but rather the volumes”

0.73

Housing 'is' the business cycle: looking at quarterly data since 1947, the housing sector signaled the oncoming recession in 8 of 10 cases by about three or four quarters, with the only two non-housing/non-consumer downturns being the 1953 'Department of Defense' disarmament downturn and the 2001 'internet comeuppance' business downturn; the two false positives (housing weakened but no recession) occurred when Department of Defense spending offset the weakness (Korean War buildup 1951, Vietnam buildup 1967).

factualhigh valuecontestednovelty 3/4durability 3/4· Ed Leamer

“in eight out of ten cases housing sector had signaled the oncoming recession given you an early warning by about three or four quarters”

0.73

The housing sector—unlike the broader economy—was genuinely afflicted by a credit crunch, because borrowers who had purchased with no money down could no longer use refinancing to pay interest and a large fraction of potential homebuyers were simply denied credit they probably should not have had in the first place; meanwhile the rise in corporate credit spreads was partly just a return to normal after a long period of underpricing risk.

causalhigh valuecontestednovelty 3/4durability 3/4· Ed Leamer

“the housing sector was very much afflicted by a credit crunch meaning our credit crunch I mean a dramatic increase in lending standards higher interest rates but most importantly just outright denial”

0.73

Rapid price formation in housing requires motivated sellers; owner-occupants are not motivated (they hold out for their price), but builders historically are because they don't want to carry inventory, and in the current crisis foreclosing banks have become the motivated sellers driving the market in an unprecedented way.

causalhigh valuecontestednovelty 3/4durability 3/4· Ed Leamer

“the key thing for rapid price formation is you need to have motivated sellers and a typical owner occupant is not a particularly motivated seller”

0.69

The US economy has grown in an extremely narrow corridor of about 3% per year for 35 years despite enormous variation in policies (oil shocks, Reagan and Bush tax cuts, the internet boom, monetary policy, the peace dividend, trade and budget deficits), which implies that the policies within the range we have experimented with have little effect on long-term economic growth.

factualhigh valuecontestednovelty 3/4durability 2/4· Ed Leamer

“the US economy has been in a very narrow corridor economic growth for 35 years growing a 3% per year sometimes a little more sometimes the last”

0.69

The Federal Reserve's extraordinarily low real interest rates in 2002-2004—maintained even as the economy and job market improved because the Fed suffered from 'deflation phobia' (treating deflation as a cause rather than a symptom and fearing a Japan-style episode)—were the 'teaser rates' that supported the subprime market, and when the Fed then sharply raised rates it killed the subprime market, dropping housing volumes and ending price appreciation.

causalhigh valuecontestednovelty 3/4durability 2/4· Ed Leamer

“big surprise that ended the subprime market it takes out a huge class of buyers how the housing market so volumes drop”

0.68

Real business cycle theory imposes a severe conceptual straightjacket on the data by attributing recessions to mysterious 'productivity shocks' whose source is never specified, while ignoring the actual facts about how the economy behaves.

factualhigh valuecontestednovelty 2/4durability 3/4· Ed Leamer

“what's wrong with that is they don't tell what the shock is and they don't look at any real facts about how the economy actually behaves”

0.68

Because long-term growth appears insensitive to policy within the experienced range, macroeconomic policy attention should be redirected away from trying to boost long-term growth and toward understanding and preventing recessions—the dips within the corridor.

normativehigh valuecontestednovelty 2/4durability 3/4· Ed Leamer

“the emphasis policy emphasis within the range that we've been dealing with ought to be worried about recessions and not long-term economic growth”

0.68

The Korean War episode offers a relatively clean fiscal experiment: a huge ramp-up in Department of Defense spending in 1951 offset a weak economy, and the large exogenous cutback in 1953 (post-armistice, defense falling from ~15% to ~8-9% of GDP) tipped the economy into recession—suggesting government spending can affect downturns at scale, though most historical episodes give only one data point each.

causalhigh valuecontestednovelty 2/4durability 3/4· Ed Leamer

“the Korean War episode is a great one because you see fiscal policy and growing in two different directions a huge ramp up in DoD spending when the economy is weak in 1951 and then in 1953 when the economy is evidently ok you see a huge cutback”

0.68

Consumer durables (especially autos) weaken just after housing both because they are interest-rate-sensitive sectors killed off together by Fed rate hikes, and because home price appreciation had made people feel able to afford more and more expensive cars—so when appreciation stops, the loss of that 'income flow' causes consumers to cut back on postponable durable purchases (repairing rather than replacing).

causalhigh valuecontestednovelty 2/4durability 3/4· Ed Leamer

“the appreciation of home prices made people think they could afford more autos and more expensive automobiles so when that stops suddenly you have a in a sense the incomes income flow becomes a lot less”

0.68

The bailouts have disastrous incentive and redistributive consequences and encourage a loss of faith in capitalism—partly correctly, because firms are exploiting the political system rather than the capitalist system, which should instead produce a loss of faith in politics.

normativehigh valuecontestednovelty 2/4durability 3/4· Russ Roberts

“the incentives for the future disastrous from that as well which really disturbs me that redistribution all aspects are atrocious it it encourages a great lack of faith in capitalism”

0.68

Government does not 'create jobs' in a normal economy—its spending takes jobs from the private sector, so the cost of government spending is the forgone private activity; stimulus only has potential when there is genuine excess capacity and unused workers, and to work it must be focused precisely on the sectors with that slack (housing, autos, restaurants, retail), whereas the $787 billion package directed only about $8 billion at those critical sectors.

causalhigh valuecontestednovelty 2/4durability 3/4· Ed Leamer

“in a normal economy when the government spends it takes jobs from the private sector doesn't create jobs so the cost is the forgone activities in the private sector”

0.68

Displaying data visually to tap into humans' strong visual-information processing skills, and forming stories informed by macro patterns, could make economics if not more scientific then at least produce better literature and be more honest.

normativehigh valuecontestednovelty 2/4durability 3/4· Ed Leamer

“I would like to elevate the role that data actually play you know I thinking by allowing us to display it visually and tap into the processing skills for visual information that we all have”

0.68

The myth-making after the Great Depression is a cautionary parallel: whatever happens, the economy will eventually recover, and if it recovers within a year the fiscal stimulus will be trumpeted as a success regardless of whether it actually caused the recovery.

forecasthigh valuecontestednovelty 2/4durability 3/4· Russ Roberts

“regardless of what happens at some point the economy is going to recover if it recovers in the next year fiscal policy will be trumpeted as great success”

0.65

Economic models have mathematical properties but also messages; the quality economist understands the difference and treats models as rhetorical devices or metaphors that help us understand how the real economy operates, rather than taking every mathematical property literally.

normativehigh valuecontestednovelty 2/4durability 4/4· Ed Leamer

“the economic models that we have have mathematical properties but they also have messages and that the quality economist is able to understand the difference and not to think of the models is literally true”

0.63

Bailing out a firm like General Motors that is hemorrhaging cash—when auto sales have collapsed from 16 million to about 9 million units—merely dissipates taxpayer money, because the only short-run hope for the sector is to bring buyers back; absent that, GM will limp along and inevitably go bankrupt anyway.

causalhigh valuecontestednovelty 2/4durability 2/4· Ed Leamer

“you do a cash infusion to your through this troubled corporation that's hemorrhaging cash that's just a recipe for a continued hemorrhaging by the taxpayers money is just going to be totally dissipated”

0.63

The Lehman bankruptcy created great fear about solvency among financial counterparties, triggering a stampede toward Treasuries that drove up the Ted spread, after which Paulson and Bernanke—seeing the spike on their watch—assembled the TARP plan and threatened a Great Depression to get it passed.

causalhigh valuecontestednovelty 2/4durability 2/4· Ed Leamer

“the Lehman bankruptcy right or wrong I create a great fear of great concern about solvency among all the counterparties in the in the financial sector and the result was there was a stampede toward Treasuries”

0.63

A large number of home purchases in the early 2000s were second homes bought on speculation as an alternative to the plummeted stock market, encouraged by the 1997 Tax Act's elimination of capital gains on homes (including second homes lived in two of the last five years), so many such houses were unoccupied and their prices behaved differently than historically.

causalhigh valuecontestednovelty 2/4durability 2/4· Russ Roberts

“an enormous number of home construction house construction and home purchases in the early part of the 2000s were second homes the people bought on spec for speculative speculative purposes”

0.62

There is little hard evidence that a credit crunch affected the real economy: non-financial corporations had healthy balance sheets with more cash than debt, and investment spending was fairly normal until Q4 2008, when it dropped not because credit was unavailable but because firms realized consumers were cutting back and a major downturn was coming.

causalhigh valuecontestednovelty 3/4durability 2/4· Ed Leamer

“the balance sheets of non-financial corporations had been very healthy at that time there's a lot of cash more cash than there was indebtedness”

0.61

Stories are remembered and processed differently—and more powerfully—than left-brain analytical information, but PhD economists are not trained to teach literature and cannot easily give exams based on stories, so they fall back on graphs and equations that are good for exams and easy to teach.

factualhigh valuecontestednovelty 2/4durability 3/4· Russ Roberts

“we're not good at teaching literature as PhD economists we're not going to train to teach literature and now it's also hard to give exams based on stories so I think a lot of economists fall back on the sort of standard graphs and equations approach”

0.61

Housing is only about four or five percent of GDP (residential investment), yet macroeconomists' reflexive reaction ('why should this one little sector ripple through the whole economy' and 'we don't do sectors') reflects a refusal to look at disaggregated data.

factualhigh valuecontestednovelty 2/4durability 3/4· Ed Leamer

“houses are pretty small fraction of GDP in residential investment only four or five percent... we don't do sectors”

0.60

The 2008 downturn was fundamentally a Main Street (consumer) problem that policymakers' response converted into a Wall Street panic: Paulson, Bernanke and Bush threatening a 'Great Depression' to pass TARP unleashed consumer fear (visible as a spike in Google searches for 'Great Depression' in mid-September and a sharp rise in the savings rate), producing the catastrophic coordinated drop in consumer spending in Q4 2008 that mechanistic theory said could not be coordinated.

causalhigh valuefringenovelty 4/4durability 2/4· Ed Leamer

“the harm that that unleashes this fear on the part of consumers that the Great Depression is coming and they all of a sudden become defensive”

0.60

It is no surprise that banks receiving federal capital infusions are not ramping up lending—we don't want them giving loans to people who aren't going to pay it back, which is exactly the bad practice that caused the problem.

normativehigh valuecontestednovelty 2/4durability 2/4· Ed Leamer

“we're totally surprised that these banks have had capital infusion from the federal government are not ramping up in lending well we don't want them to give loans to people aren't going to pay it back”

0.60

We will never know what would have happened if the Fed had let Bear Stearns fail in March 2008; the claim that Bear's millions of counterparty contracts made its bankruptcy uniquely catastrophic is undercut by the fact that Lehman Brothers did go through bankruptcy, suggesting it might have been better to let the pain be short and intense.

forecasthigh valuecontestednovelty 2/4durability 2/4· Russ Roberts

“the question we'll never know the answer to is what would have happened if the Fed had let Bear Stearns fail in March of oh eight”

0.56

Stable, semi-constant monetary policy explains why the economy has been so stable for the last thirty-plus years; this Friedman-style argument was a good argument until about a year and a half ago but may not pan out over the next few years.

causalhigh valuecontestednovelty 2/4durability 2/4· Russ Roberts

“we know why the economy's been so stable for the last thirty plus years it's because monetary policy has been very stable and very steady”

0.35

Forecast uncertainty grows the further out one predicts, so a one-or-two-quarter-ahead real GDP forecast carries error bands of roughly plus or minus two percent while a ten-year forecast might carry plus or minus thirty percent.

factualestablishednovelty 1/4durability 3/4· Ed Leamer

“if I was predicting a quarter or two ahead Arab and as only Warner is plus or minus two percent”

0.30

9.8 m/s² doesn't have the validity for the economy that it has for a physical system, yet it is remarkable that despite all the policy changes the US economy has kept growing at about 3% per year.

factualestablishednovelty 1/4durability 2/4· Ed Leamer

“it isn't in 9.8 m/s^2 doesn't have the validity of a constant of a physical system but it's still rather remarkable that they for all these things that US economy has been growing at 3% per year”

0.12

Macroeconomics is a discipline.

factual· Ed Leamer

“we ought to have a totally different kind of standard”