What this covers

This is a conversation between Russ Roberts and Steve Fazzari of Washington University in St. Louis, held while stimulus plans were being debated in early 2009. Roberts presses the neoclassical objections while Fazzari builds the Keynesian case from everyday examples: a family that eats out less to save for college, a restaurant that loses income, a bank that never receives the money. Robinson Crusoe's island, where forgoing bananas directly funds planting seeds, serves as the foil for how a monetary economy differs.

The middle of the conversation covers a lot of ground. Fazzari explains why the paradox of thrift is hard to see from inside one's own bank balance. Roberts answers with creative destruction: the rise of home cooking shrank restaurants and grew grocery stores without catastrophe. Fazzari argues that wage and price adjustment is no safe fix, since deflation raises the real burden of debts like a $100,000 mortgage. On policy, he sees no meaningful difference between a rebate, a helicopter drop, or paying workers to dig and refill holes. He also argues that with idle resources, stimulus need not be inflationary, which puts him at odds with Milton Friedman and simple monetarism.

The sharpest turns come on evidence and risk. Fazzari concedes that macro evidence is largely anecdotal, because history cannot be rerun, and the Reagan tax cuts and the 1982 unemployment peak are discussed in that light. His own larger worries are that stimulus money goes to less productive projects, that rent-seeking harms the investment climate, and that lingering fear, an animal spirits concern, could blunt any package.

Sharpest takeaway

Fazzari argues that in a monetary economy saving cannot raise total investment in the short run, because saving destroys income elsewhere, so recessions need a demand-restoring response that wage and price flexibility cannot reliably supply.

  • Higher saving by one group cuts another group's income, so aggregate saving does not rise (paradox of thrift).
  • Wage and price adjustment can be deflationary and hurt debtors, so it is not an automatic route back to full employment.
  • Stimulus works through demand and can raise real output when resources are idle, though its effects are hard to verify empirically.

The claims · ranked59 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.

0.86

Adjustment by price cuts and wage cuts can fail to restore demand, and deflation is harmful because it raises the real burden of nominal debts such as a $100,000 mortgage, causing defaults and tighter spending.

causalhigh valueestablishednovelty 3/4durability 4/4· Steve Fazzari

“if you have a hundred thousand dollar mortgage and your wages are going down it's going to be harder to pay your mortgage and you might default”

0.86

When a family saves by eating out less, the restaurant's lower income forces it to draw down its own savings or cut its own spending, so the net increase in household saving is exactly offset and aggregate saving does not rise.

causalhigh valueestablishednovelty 3/4durability 4/4· Steve Fazzari

“the family that decides to eat at home rather than eating out to saving more but by exactly the same amount the restaurant owner and its employees are saving less”

0.86

Macroeconomic evidence for or against stimulus is largely anecdotal: economists have only a few historical episodes, cannot rerun history with different policies, and so cannot settle which view is correct, and many macro events are historically unique single observations.

causalhigh valueestablishednovelty 3/4durability 4/4· Steve Fazzari

“we just have a few historical periods in which these kinds of things were were attempted we can't rerun those histories with different policies”

0.80

Keynes's animal spirits argument holds that weak expectations and fear about the future can lead to a depressed economy, a very Keynesian argument, and the investment-climate concern follows from it.

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

“you made a very good Keynesian argument in some sense about what what Keynes called animal spirits that that weak expectations and fear about the future can lead to a depressed economy”

0.80

The paradox of thrift is hard to grasp because when an individual saves more, the individual sees rising bank balances and portfolio values, but does not see the income destroyed elsewhere in the economy by the reduced spending.

causalhigh valueestablishednovelty 2/4durability 4/4· Steve Fazzari

“when we individually save more we see that our our bank accounts rise our stock portfolio increases ... we do not see the income that we've destroyed elsewhere in the economy”

0.78

In a monetary economy, a decision to save cannot turn into investment automatically: unlike Robinson Crusoe's barter-like island, where forgoing bananas to plant seeds directly creates future output, saving in a monetary system is not seen by the market as a signal for investment.

causalhigh valuecontestednovelty 3/4durability 4/4· Steve Fazzari

“the difference between a monetary economy and an economy where all trade is in terms of goods and services... is really very important to this story”

0.78

When a household reduces restaurant spending to save for its children's college, the market sees the reduced restaurant demand and adjusts by cutting restaurant food, but does not see the decision to fund college; the bank never gets the money, so the intended reallocation does not happen.

causalhigh valuecontestednovelty 3/4durability 4/4· Steve Fazzari

“the market system sees that first effect they see that I'm not going on to eat is often and will adjust to it... but they will not see my decision to send my kids to college”

0.76

If everyone stopped spending and saved all income, total income would fall to zero, because one person's spending is another's income; this implication is standard in mainstream textbooks that discuss short-run declines in aggregate demand.

causalhigh valueestablishednovelty 2/4durability 4/4· Steve Fazzari

“think about what would happen if people every people decided to save everything today... income would fall to 0”

0.75

The helicopter drop works only if recipients fail to anticipate it, so repeated money creation would be recognized and lead to inflation rather than real output growth; the key is expectations.

causalhigh valueestablishednovelty 2/4durability 3/4· Steve Fazzari

“I think the issue there was a question of expectations right if people were aware”

0.74

Holding total income constant while saving rises is a logical fallacy: an increase in saving necessarily reduces total income, so there is no additional money available to finance investment unless some other process restores demand.

causalhigh valuecontestednovelty 3/4durability 4/4· Steve Fazzari

“when people save more we cannot keep income constant so an increase in saving necessarily destroys income so there is no additional money in the sense of income flows over time to invest”

0.74

Critics of the Keynesian view respond that a broad shift in spending, such as the 1980s-2000s rise of home cooking, does not cause massive dislocation, because creative destruction reallocates resources: restaurants shrink while grocery stores expand and prices steer resources to their highest-valued uses.

causalhigh valueestablishednovelty 1/4durability 4/4· Steve Fazzari

“that kind of creative destruction is going on all the time in a market economy it's not coordinated the prices are adjusting to steer the resources to their highest valued uses”

0.73

Saving is not automatically the engine of growth: when a family saves by cutting restaurant spending, the restaurant's income falls, and a decision not to consume today is not necessarily a decision to consume something specific later.

causalhigh valuecontestednovelty 3/4durability 3/4· Steve Fazzari

“a decision not to have dinner today is not necessarily a decision to have dinner tomorrow or to buy a pair of boots tomorrow”

0.73

Government printing of money and borrowed government spending have the same effect on demand as a private-sector tax rebate, because each puts money in people's hands; Fazzari says he sees no meaningful difference between a rebate, a helicopter drop, or hiring workers to dig and refill holes.

causalhigh valuecontestednovelty 3/4durability 3/4· Steve Fazzari

“I'm not sure I see any difference between the government printing money and giving it to people 160 billion worth and saying oh I hope they spend it”

0.73

In a world of unemployed resources, a monetary or fiscal stimulus need not be inflationary, because higher demand can raise real production by hiring idle workers rather than raising prices; only in the long run, when resources are fully employed, would the stimulus be inflationary.

causalhigh valuecontestednovelty 3/4durability 3/4· Steve Fazzari

“a rise in the supply of money can lead to higher demand higher demand will in some circumstances lead to higher prices but higher demand can also lead to higher real production”

0.73

A simple monetarist view links the money supply automatically to the price level, but Fazzari does not see such an automatic link; the link works through demand, so when there are unused resources an increase in money raises real output rather than prices.

causalhigh valuecontestednovelty 3/4durability 3/4· Steve Fazzari

“I don't see that as an automatic link I think it works through demand basically so a rise in the supply of money can lead to higher demand”

0.70

Falling wages do not automatically raise demand, because a lower wage bill does not create a channel that restores spending; the standard story's price-cutting channel is subtle and not guaranteed to fill the gap created by higher saving.

causalhigh valuecontestednovelty 3/4durability 3/4· Steve Fazzari

“why should these workers a good jobs i if businesses are not going to hire workers to produce stuff that they can't sell so somehow there has to be a channel through which lower wages will lead to higher demand”

0.69

A policy that lowers the consumption-to-income ratio can lead to higher growth only if resources are reallocated toward productive investment while aggregate demand is maintained at its prior level; both must happen, and from the Keynesian perspective maintaining adequate demand is not automatic.

causalhigh valuecontestednovelty 2/4durability 4/4· Steve Fazzari

“those two things together will lead to higher growth the problem is that they both have to happen and from the kansian perspective keeping aggregate demand at an adequate level is not automatic”

0.69

The Reagan tax cuts of the early 1980s were permanent and expected to persist; although intended to stimulate saving, much of the effect seemed to go to consumption, and the economy did reasonably well with relatively low tax rates through the 1990s and much of the 2000s, even though unemployment reached 10.8 percent in late 1982.

factualhigh valuecontestednovelty 3/4durability 2/4· Steve Fazzari

“unemployment rate since the depression a 10.8% late 1982 ... massive tax cuts in this case permanent tax cuts that people would not expect to be rescinded”

0.68

A coordinated society-wide decision to work and consume less could make everyone better off, but in a decentralized economy no such coordination occurs, so a reduction in spending by one group leaves businesses selling less and forces layoffs.

causalhigh valuecontestednovelty 2/4durability 3/4· Steve Fazzari

“any individual group of individuals that had been spending a lot decide to cut back they may be better off they may see themselves being better off but the businesses that we're selling them goods and services are not selling as much”

0.68

Milton Friedman argued that a demand-driven stimulus by hiring workers would have only a temporary real effect and would mainly raise prices, because people would eventually realize there was no real increase in demand for their services and return to the original equilibrium.

causalhigh valueestablishednovelty 2/4durability 3/4· Steve Fazzari

“what Melton Friedman argued I think is that that effect would be temporary that the net effect would be merely to raise prices”

0.68

The Fazzari-backed concern is that a large stimulus might still fail if people remain fearful about the future, because uncertainty over bailouts and the rules of the game leads households and banks to wait on the sidelines and cut back spending regardless of the stimulus.

causalhigh valuecontestednovelty 2/4durability 3/4· Steve Fazzari

“people are fearful of the future right part of the reason they're fearful is they just think the economy is gonna go down part of the reason they're fearful of the future is that they're not sure what the rules of the game are going to be”

0.68

If resources shift from consumption today to investment tomorrow, aggregate demand should be unchanged, because the restaurant contracts while construction firms expand as more funds become available to spend on investment.

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

“isn't that simply going to keep aggregate demand unchanged... isn't the construction firm going to be expanding”

0.68

Most modern neoclassical growth models assume adequate aggregate demand, because they assume wage and price adjustment eventually restores it; Keynesians like Fazzari do not consider that process automatic even over a long horizon.

causalhigh valuecontestednovelty 2/4durability 3/4· Steve Fazzari

“the most modern growth models neoclassical growth theory assumes there will be adequate aggregate demand why because they assumed it ultimately that wages price adjustment process”

0.68

The bigger worry about a large stimulus package is not that it fails to stimulate but that its spending goes into less productive projects than private-sector alternatives, and that it worsens the investment climate through rent-seeking.

normativehigh valuecontestednovelty 2/4durability 3/4· Steve Fazzari

“what I worry about more is the investment climate which is people are fearful of the future”

0.66

If saving is the engine of future growth because saved funds finance investment, then saving should be good for the economy, and Roberts does not understand why higher saving would be harmful.

causalhigh valueestablishednovelty 1/4durability 4/4· Russ Roberts

“surely savings and therefore investment are the engine of future growth so savings is good so why is it that savings is bad”

0.65

The standard mainstream account is that when saving rises, unemployed resources and falling wages push prices down throughout the economy, stimulating demand until resources are re-employed; this adjustment period is the transition between the short run and the long run.

causalhigh valueestablishednovelty 1/4durability 3/4· Steve Fazzari

“the standard story is something like this that there are unemployed resources as the result of this falling demand in particular unemployed labor”

0.64

In normal times, shifts in spending across sectors are reallocations that do not hurt the aggregate economy, but a systemic, broad-based drop in consumption across all sectors, caused by households losing access to credit, is a net decline in total demand.

causalhigh valuecontestednovelty 2/4durability 3/4· Steve Fazzari

“it's not just a matter of reallocating goods and services from one sector to another it's a net drop in total demand and a significant one throughout the economy”

0.64

Shifting the tax system to treat consumption and saving neutrally, or to favor saving, would raise saving today and lower output today, but raise output in the future as the saving turns into investment and innovation.

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

“wouldn't that yield a higher savings rate today perhaps a lower rate of output today as a result of that but a higher output in the future as that savings turns into investment”

0.64

Fazzari argues that recent US economic growth was supported by an autonomous drive toward more demand and a lending-based financial system, but that this process is not automatic and can falter under particular historical circumstances, as it has now.

causalhigh valuecontestednovelty 2/4durability 3/4· Steve Fazzari

“there has been I might think of as a kind of an autonomous drive to more demand and it's and we've also had a financial system that's facilitated by by lending more”

0.63

Many economists, Fazzari included, now believe demand has fallen so far that private demand cannot recover on its own, so the government must exogenously raise demand through spending, which it can do through borrowing or printing money because of its credit.

normativehigh valuecontestednovelty 2/4durability 2/4· Steve Fazzari

“many economists myself included now believe that the problems of demand are so severe that we cannot rely on any of the private sources of demand to rise on their own”

0.61

The point of government spending in a demand-shortfall is not which projects are chosen, such as roads, hospitals, or bridges, but that total demand rises; the reversal of the paradox of thrift is that more spending creates the income that finances it.

causalhigh valuecontestednovelty 2/4durability 3/4· Steve Fazzari

“the basic point is less about whether we need more roads or more hospitals or more bridges or more of anything in particular but that we need more demand in total”

0.61

Robert Higgs argued that wartime prosperity was an illusion because wage and price controls distorted measured activity, so the war reduced unemployment through conscription and military production without stimulating the broader economy; Fazzari disagrees with that view.

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

“our podcast with Robert Higgs argued that the wartime prosperity was an illusion that measurement of economic activity is very difficult during the war due to wage and price controls”

0.56

World War II, Korea, and Vietnam-era spending periods show that government stimulus can raise growth and reduce unemployment: unemployment fell to near zero during World War II and the large deficits were associated with fast growth.

factualhigh valuecontestednovelty 2/4durability 2/4· Steve Fazzari

“unemployment fell to virtually zero lots of incomes were created there was a huge government deficit so the economy was stimulated”

0.55

A simple cash transfer such as a tax rebate is not stimulative by itself: the effect depends on people spending the money, and a recent estimate suggests only twenty to fifty percent of the 2008 rebate was spent.

factualhigh valuecontestednovelty 2/4durability 1/4· Steve Fazzari

“maybe twenty to fifty percent was was spent so much less than much much less than a total that was put out there”

0.52

A society that voluntarily chose to work and consume less, valuing leisure and family, could be better off in real utility and happiness even though measured GDP would be lower, so consumption has no inherent value in that case.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Russ Roberts

“measure GDP would be lower but real utility happiness joy satisfaction could be dramatically higher”

0.52

A widespread shift toward a less materialistic, less work-intensive lifestyle would raise utility and well-being even though measured GDP would fall, but such a shift would be painful to adjust to if it happened quickly and uncoordinated.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Russ Roberts

“measure GDP would be lower but real Genie low real whatever you want to call utility happiness joy satisfaction could be dramatically higher”

0.51

Macroeconomic errors come largely from neglecting to allow something to change that has to change, such as treating savings as an exogenous shock when savings are usually caused by other factors like tax policy.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Steve Fazzari

“inevitably I find that most macroeconomic errors in thinking ... comes from neglecting to allow something to change that has to change”

0.44

When a family cuts back restaurant meals to cook at home, the restaurant's revenue and sales fall and it has a harder time generating the business it is used to, which raises the question of what adjustment the economy must make.

causalestablishednovelty 1/4durability 4/4· Steve Fazzari

“the restaurants they were going to are going to see their revenues fall their sales fall”

0.39

When many people cut back on spending at once, as in the paradox of thrift, the adjustment is painful in the short run and dependent on how quickly it happens; a gradual change over ten years would be far less disruptive than an overnight change.

causalestablishednovelty 1/4durability 3/4· Steve Fazzari

“if over ten years we decided to slowly reduce our material spending and slowly reduce our amount of work I think that would be a different story”

0.35

If the government printed money to employ construction workers, those workers would treat the wages as income, and previously idle resources would be brought into use, creating real income through higher demand.

causalestablishednovelty 1/4durability 3/4· Steve Fazzari

“those construction workers are going to think of that is income... those resources that were idle are brought into use”

0.35

Most mainstream macroeconomists, following Friedman, hold that in the long run the economy fully employs its resources, so stimulus whether fiscal or monetary is inflationary unless offset once the economy recovers, a view Fazzari partly rejects.

forecastestablishednovelty 1/4durability 3/4· Steve Fazzari

“in the long run we will fully employ our resources so in the long run whether it's fiscal or monetary stimulus that would be inflationary”

0.35

The inflationary impact of stimulus is a real worry, and it would arise regardless of whether the money is delivered as checks or through hiring workers, if the stimulus actually works.

forecastestablishednovelty 1/4durability 3/4· Russ Roberts

“what I would worry about there is is the inflationary impact of that and I think that impact is going to be there regardless”

0.34

The move from a short-run saving-induced income loss to a long-run higher saving rate depends on wage and price adjustment, which has been a core debate in macroeconomics for roughly seventy years since Keynes.

factualestablishednovelty 0/4durability 4/4· Steve Fazzari

“that has to do with some very subtle issues that have been a core of the debate in macroeconomics I think for at least since Keynes that we're talking about seventy years plus about wage and price adjustment”

0.34

Monetary policy alone may not be enough because the Fed's wage-and-price-adjustment role requires a Taylor rule or inflation targeting, and a wage-price deflation spiral is a risk that an enlightened monetary policy must counter.

normativeestablishednovelty 1/4durability 2/4· Steve Fazzari

“so an enlightened monetary policy should step in to fill the role of a wage and price adjustment hence the idea of the things you might have talked about in your past broadcasts the Taylor rule or or inflation targeting”

0.34

Under the Keynesian view, the most direct determinant of output, production, and employment is aggregate demand, meaning the amount of money people are spending in the economy, so a drop in aggregate demand causes a recession.

causalestablishednovelty 0/4durability 4/4· Steve Fazzari

“the most direct determinant of output production employment in the Keynesian view is the amount of money people are spending in the economy what economists call aggregate demand”

0.32

Roberts counters that the steel makers and software firms supplying an investment boom would have been funded by the saved money, so their activity offsets the restaurant's loss of income and total income is not necessarily reduced.

causalcontestednovelty 1/4durability 3/4· Russ Roberts

“why isn't that offset by the person who is selling the materials that are going to be used for the investment”

0.30

Economists should respond to the difficulty of confirming theories by taking evidence seriously, relying more on case studies and economic history, and staying open-minded about alternative theories rather than confirming priors.

normativespeaker onlynovelty 1/4durability 3/4· Steve Fazzari

“economics has tended focus on formal statistical methods that require lots of data maybe we need to look more at case studies and economic history”

0.29

The current downturn is driven by a reduction in consumer spending because households borrowed against their houses to finance consumption and can no longer do so, so a drop in home-related borrowing cuts business sales and hiring.

causalcontestednovelty 1/4durability 1/4· Steve Fazzari

“people were borrowing lots of money against their houses to finance their consumption spending so what they don't use their house as a piggy bank anymore they won't spend as much”

0.27

The government can finance a demand-raising spending package without a financing constraint, because it can print money or borrow given its good credit, and the package needs to be pretty significant.

normativecontestednovelty 1/4durability 2/4· Steve Fazzari

“the government does not have to because effectively because they can print money or or borrow so far it's good credit”

0.24

Economic policymakers such as Paulson, Bernanke, and Obama face a very difficult job and cannot debate Keynesian versus non-Keynesian theory before acting, so policy will involve starting and stopping and some loss of confidence.

normativespeaker onlynovelty 1/4durability 2/4· Steve Fazzari

“they can't afford to debate about the world Keynesian or non Keynesian they have to come up with policies that affect people's lives now”

0.23

The current US consumption decline is not a voluntary choice for less materialism but reflects households being financially overstretched, so workers who want jobs cannot find them.

factualcontestednovelty 1/4durability 1/4· Steve Fazzari

“what we're facing right now in reality with the problems the US economy is going through at the moment is a decision to consume less not because people felt like they should take leisure but because they were financially overstretched”

0.21

The empirical question is how quickly wages and prices adjust to remove the aggregate demand problem; mainstream middle-of-the-road Keynesians thought adjustment was too slow to wait for, and this view changed only very recently.

factualspeaker onlynovelty 1/4durability 2/4· Steve Fazzari

“the question is how quickly will wages and prices adjust so that the aggregate demand problem goes away”

0.21

Keynesian macroeconomics' argument about saving is one of the most subtle issues in the field, and most of the economics profession, perhaps most of it, misses it.

normativespeaker onlynovelty 1/4durability 2/4· Steve Fazzari

“many of the economics profession maybe most of the economics profession is missing something”

0.18

Fazzari expects the proposed stimulus package to make the economy worse through rent-seeking and to fail to change expectations, and he sees no empirical evidence that it will.

forecastspeaker onlynovelty 1/4durability 1/4· Steve Fazzari

“if anything I think it's going to make it worse because of rent-seeking I don't see any evidence that this stimulus package is going to change people's expectations”

0.18

Fazzari has concluded after a quarter century of research that the Keynesian perspective is the best way to understand the macro economy, while trying to remain open to alternatives.

normativespeaker onlynovelty 1/4durability 1/4· Steve Fazzari

“I have come to conclusion over a quarter century plus of research and teaching on these areas that the Keynesian perspective is the best way to understand the macro economy”

0.18

The current trillion-dollar stimulus will be a near-natural experiment that teaches economists something, though the results will likely be debated.

forecastspeaker onlynovelty 1/4durability 1/4· Steve Fazzari

“if we have a trillion-dollar stimulus package which some people are talking about we are going to learn something I think we may debate it and argue about what it actually means”

0.18

Monetary policy alone may not be enough to counter the current downturn, so other measures might be needed.

forecastspeaker onlynovelty 1/4durability 1/4· Steve Fazzari

“the concern is monetary policy alone won't be enough so we might need something else”

0.17

The current recession reflects insufficient spending: consumer spending and residential construction are falling, net exports are weakening as foreign economies slow, and the Fed cutting interest rates to near zero may not be enough to restore demand.

factualcontestednovelty 0/4durability 1/4· Steve Fazzari

“we're facing a problem of insufficient spending ... consumer spending is declining residential construction is it a terrible slump net exports”

0.10

Thanks for listening and see you next Monday.

factualspeaker onlynovelty 0/4durability 0/4· Russ Roberts

“talk to you on Monday”