Scott Sumner  

Lake Wobegon Keynesians

The Case Against Education<... The Devastating $15 Minimum Wa...

The Keynesian view of fiscal policy is an ever-changing theory. In the 1960s, it was believed that fiscal stimulus was needed during recessions. In the 1990s, Keynesians argued that fiscal policy was not an appropriate stabilization tool, rather monetary policy should be employed. Then after 2008, Keynesians suggested that fiscal policy was OK, but only when the economy was at the zero bound. And now even that qualifier is being gradually phased out.

Keynesians also insist that the stance of fiscal policy depends on the change in the cyclically adjusted deficit, not the level of the deficit in absolute terms. In other words, it would be impossible to run a continually expansionary fiscal policy. For every year of expansionary fiscal policy, there will be another year of contractionary fiscal policy.

If the Keynesians are right that it's the change in the deficit that matters, then you'd obviously want to do fiscal stimulus when unemployment is high, and fiscal austerity when unemployment is low. So is that what Keynesians actually do recommend? Apparently not---here's the Economist:

With total revenues of C$300 billion, the federal government's capacity to lift the C$2 trillion economy is limited. The budget's impact will be modest, says Craig Alexander of the C.D. Howe Institute, a think-tank. Still, most economists support deficit spending at a time when borrowing rates are low and the economy is weak. The question is: will Mr Trudeau know when to stop?
As far as "when borrowing rates are low", that would be the next 100 years, wouldn't it? Are we to have fiscal stimulus for the entire 21st century? The Keynesians say that's impossible, what matters is whether the deficit is bigger or smaller than normal. So how about the timing? Here's the unemployment rate in Canada:

Screen Shot 2016-03-31 at 11.10.28 AM.png
It ticked up slightly in recent months, due to a slow patch in the first half of 2015. But Canada is expected to grow again this year. If no recession, then why do fiscal stimulus? And "will Mr Trudeau know when to stop?" Let me guess, right before the next big recession.

And why not cut interest rates instead? There are no good answers from Keynesians, just mumbles about the "need" for infrastructure, or the worry about "bubbles". Yes, but what does any of that have to do with countercyclical fiscal policy? They have no answers.

Canada has a done a better job than most countries in holding down its public debt. And now it's going to just throw those hard-earned gains away, for no good reason at all.

Screen Shot 2016-03-31 at 11.28.32 AM.png
Some Keynesians seem to be like those deluded residents of Lake Wobegon, believing that fiscal policy can always be more expansionary than average.

And if you think things are bad in Canada, consider that in America many Keynesians favor fiscal stimulus, even with unemployment at 4.9%. What will they advocate when the next recession hits? Even more fiscal stimulus?

Comments and Sharing

COMMENTS (14 to date)
jc writes:

When I'm in the mood, I ask my most outspoken Keynesian friends if they're actually Keynesian, since their preferred policies over the years correlate almost perfectly w/ the view that Keynes' main ideas simply provide them w/ an intellectual veneer for their true goal of increasing the size and scope of the state.

Counter-cyclical fiscal policy? Please. They never support, or even mention it, when times are good.

(To be clear, I'm not saying that non-Keynesians are immune to similar practices that support their own true beliefs. We're all human, after all, and this is simply how we're wired. And, fwiw, I think critical analysis of, say, more liberty minded folks is a good thing too...anything that imposes analytical discipline and forces folks of any persuasion to safeguard against the all-too-easy inclination to use our grey matter to justify and rationalize rather than analyze is good and healthy. I'd actually like to read more of this too, i.e., posts/comments aimed at genuinely sharpening pro-liberty arguments.)

Back on friends do make concessions in the form of theory modification during times when the academic winds and evidence have blown a bit more in the other direction. But generally speaking the rule is simple: be as big a supporter of fiscal spending as the times will credibly allow.

I tell them I - a relatively agnostic, critical admirer of Keynes - am actually far more Keynesian than they, and most other self-declared Keynesians, are.

Sometimes I get logically coherent counters to my playful barb, justifications for seeming or actual departures from orthodox Keynesian prescription that actually do seem, well, plausibly justified in a way that both makes sense and suspends my doubt about the sincerity of their Keynesian credentials. Often I get something else.

Mark Brophy writes:

The unemployment rate is not 4.9%; that's a fraudulent government number. If you want the real story, look at the labor force participation rate, showing that the economy has been in recession for 8 years.

ThaomasH writes:

I think the shifting "Keynesian" position you observe is because of the deficit/investment implications of a) whether governments are likely to approximately follow an NPV rule and b) whether monetary authorities are likely to approximately follow an NGDP trend target rule. When it appears that both go off the reservation at once, one argues for "fiscal" policy and "unconventional" monetary policy (anything different from purchasing ST government paper).

It's not that hard.

LK Beland writes:

Just a note: these debt numbers for Canada ignore provincial debt, which is more important in many provinces than federal debt, as a fraction of GDP.

Daniel Kuehn writes:

Monetarism is an ever-changing theory. In 1898 it was about a stable price level. In the 1970s it was about a stable growth rate of the money supply. Today it's about NGDPLT.

Most of these shifts in theory you identify (especially he recent once) with Keynesianism have to do with the gradual grappling with and incorporation of expectations, which as far as I can tell is quite comparable to the scientific forces pushing monetarism through its evolution.

I am a little confused by your read of policy proposals lately. So yes, you see some calls for fiscal policy around but they are nothing like the calls for it in 2009, 2010, 2011, etc. I would have thought exactly the opposite of your claim here - I would have thought that some kind of clever social media/big data metric of demand for fiscal stimulus would correlate very well with the unemployment rate. Do you not think so?

So yes, most people haven't taken up a slash-and-burn approach to the budget but that doesn't seem all that problematic to me. We're in an unusual low-inflation, stagnant wage environment where the employment to population ratio suggests the unemployment rate may be a misleading metric (though some of the EP ratio is structural). That makes he claim that fiscal policy is done for now a little less clear to me than it is to you. But beyond that most of the calls for major spending are NOT being justified on cyclical grounds at all (Medicare for all, college for all, XYZ for all). Those are important to talk about and argue against but not really related to how we want to approach demand management.

So I would have thought we'd expect to see much less calls for fiscal stimulus for cyclical reasons but not necessarily less calls from all quarters for spending, period.

Isn't that exactly what we're seeing?

Scott Sumner writes:

Mark, If the economy has been in recession for 8 years, why has RGDP been growing for the past 6.5 years? Isn't it more likely that the LFPR is falling due to boomers retiring?

Thoamas? Why would fiscal policy be able to make up for monetary policy failing to do its job? Wouldn't the BOC merely offset any fiscal stimulus? Canada is not at the zero bound.

LK, Good point.

Daniel, Monetarism has evolved in terms of the preferred target, but not the basic idea that monetary policy does and should drive nominal variables. It seems to me that changes in Keynesianism are more basic. I have no idea why the Canadians thought fiscal stimulus was a bad idea in the early 1990s, but a good idea today.

I don't see how expectations relate to the changes I described above.

As far as the prevalence of calls for fiscal stimulus, I see them all the time, in major respected publications. And of course the article I cite says the same.

Daublin writes:

With apologies for furthering a subthread, I'm curious why unemployment gets so much play, myself. It would be a great topic for an article if you ever want to post about more basic things, Scott!

If you are trying to understand the labor market, it seems like labor force participation is a much more straightforward variable to reason about. Yes, it swings around much more wildly than unemployment, but if that's what the labor market is doing, why hide it? It seems like throwing away useful data to smush unemployment back toward 0 automatically, no matter what the underlying data are doing.

On the main topic, yeah.... the case for stimulus involves it being temporary, doesn't it? If you just stimulate continuously, it's much like treating someone with heart problems by turning the defibrillator to max and just leaving it turned on.

Benjamin Cole writes:

Nice post.

Perhaps some observers are suggesting fiscal stimulus as our central banks seem more interested in monetary asphyxiation than in growth.

If fiscal stimulus is obtained through tax cuts, and central banks stay on their inflation target, then I suspect fiscal stimulus would be worthy.

Rob writes:

This is from the Reserve Bank of Australia's statement following its decision to cut interest rates on 2 December 2008:

"There has now been a major easing in monetary policy over the past few months. Together with the spending measures announced by the Government, and a large fall in the Australian dollar exchange rate, significant policy stimulus will be supporting demand over the year ahead. The Board will continue to monitor developments and make adjustments as needed to promote sustainable growth consistent with achieving the 2–3 per cent inflation target over time."

Here's the link to the full statement:

Isn't this a fairly conventional view of policy around that time, that monetary and fiscal policy were working together to support demand, and it was useful to have both working together because they supported demand in somewhat different ways? Scott, would you say that if not for the Australian Government's spending measures around that time, the RBA would have cut interest rates even further, and this would have been just as effective in supporting demand?

Brian Donohue writes:

Good post. Important. Scott Adams once said that people do what they want, then they come up with reasons.

At bottom, people who espouse Keynesianism really just want more government spending. Always. That's the only thing they are consistent about.

Robert Thorpe writes:

Daniel Kuehn,

"Monetarism is an ever-changing theory. In 1898 it was about a stable price level."

In 1898 there was no such thing as Monetarism. The term arose much later, as you know. We now call Fisher, Cassel, etc, Pre-Monetarists retrospectively. The problem with the label "Keynesian" is that the speaker is trying to confer on himself (or herself) the authority of Keynes. Then they attach that to all manner of different theories.

Scott and the NGDP targetting crowd also adopt the label Monetarist, e.g. Market Monetarist. Whether all the other Monetarists agree with that is another matter. It's a way of assuming some of Friedman's authority. In this case has only been done once though and only by a few people. There isn't a history of doing it as there is with Keynesianism.

"Keynesianism have to do with the gradual grappling with and incorporation of expectations"

I think Krugman has been arguing for the opposite. He has defended old-style Keynesianism, quite explicitly. Many of the Keynesians I hear about these days are "New-Old" Keynesians. I think you are Daniel, to a degree.

"... employment to population ratio suggests the unemployment rate may be a misleading metric"

In Britain the employment-to-population-ratio has recovered. However, Keynesians are still trying to cook up ways of claiming that there isn't a real recovery yet.

Scott Sumner writes:

Daublin, It's hard to know what to make of labor force participation. For instance, the LFPR was lower in the 1960s than today, and yet most people think the labor market was pretty good in the 1960s. But all macro variables have issues.

Rob, You asked:

"Scott, would you say that if not for the Australian Government's spending measures around that time, the RBA would have cut interest rates even further, and this would have been just as effective in supporting demand?"

Yes, that's my view. I would distinguish between elite opinion like Paul Krugman, and average opinion. Average Keynesians thought fiscal stimulus could play a useful role even when rates were positive. Elite opinion, such as Paul Krugman, thought it was only useful when rates were zero.

Brian, I would add that both Keynesians and anti-Keynesians think Keynesian theory has big government implications, but you could also get stimulus via tax cuts.

Brian Donohue writes:

Yeah Scott, that's what I'm driving at. Stimulus 'could' be tax cuts, austerity 'could' be spending cuts, but these are not the 'go to' policies for any so-called Keynesian I know.

Which is why I think these people are big government wolves in Keynesian sheep clothing.

Ron writes:

Keynesian or monetarist policies cannot solve this problem. The USA, Europe, and Japan are in a productivity driven liquidity trap.
Every year productivity increases allow large companies to maintain or increase production and at the same time lay off employees. Productivity increases are inherently deflationary.
Increasing wages at the same rate as productivity growth counteracts this. If wage increases are equal to productivity increases it has zero impact on inflation but it does cause economic growth and it provides incentives for investment to provide future productivity enhancements.
Keynesian or monetary stimulus cannot resolve a productivity driven liquidity trap. Increased investment makes the problem worse because it further enhances productivity.
The best, and perhaps the only solution is to shift to a new economic model. "Productivity Equilibrium Economics" regulates the minimum wage to maintain equilibrium between productivity driven supply expansion and wage driven consumption (demand) expansion. Increasing demand causes economic growth and government revenue growth which provides funds for Keynesian stimulus without increasing debt leverage ratios and it provides and incentive for investment to provide future productivity and demand growth. The model incentivizes productivity increases which enhance the nation's wealth creation and competitive position.
Interestingly the recent minimum wage increases in Washington, California and New York risk the stagflation if they are not properly coupled to productivity. To learn more about "Productivity Equilibrium Economics" see

Comments for this entry have been closed
Return to top