StanCollender'sCapitalGainsandGames Washington, Wall Street and Everything in Between



I Hate To Pile On To Brian Riedl, But He Deserves It

19 Feb 2010
Posted by Stan Collender

I hate to pile on to Brian Riedl after both Brad DeLong and Matthew Yglesias do a pretty good job debunking his latest in National Review Online.

But...

There's much in Brian's piece that requires criticism, but here's the graph that is the most offensive:

The idea that government spending creates jobs makes sense only if you never ask where the government got the money. It didn’t fall from the sky. The only way Congress can inject spending into the economy is by first taxing or borrowing it out of the economy. No new demand is created; it’s a zero-sum transfer of existing demand.

Brian...The goal of economic stimulus is to create activity that wouldn't otherwise occur at that time.  Your statement would have been correct if the economy had been operating at close to full employment and capacity.  But it wasn't.  Businesses weren't spending, consumers weren't spending, and monetary policy adjustments were not doing much to change that behavior. 

So the federal government used borrowed funds (Nice try, Brian, but there were no tax increases), that is, it got funds from those who were not spending the money and then spent it or gave it to others with a tax cut (about 40 percent of the stimulus was tax cuts) to create that activity.  Unless there are no limits on productivity improvement, that has to result in more jobs than would have otherwise existed.

I'm tired of this weak argument

"that has to result in more jobs than would have otherwise existed"

Sure enough. How many more? The high numbers 1.5 to 2.0 million are the simple application of multipliers to the problem. They are the same type of analysis that would suggest if I had tripled the stimulus, unemployment would now be at 6 percent instead of 10. That doesn't mean they are wrong but it certainly doesn't mean they are right. The "bottom up" method, which I applaud the administration for trying, clearly didn't work so how many jobs is it exactly.

IMO, the strong formulation on both sides (either created no jobs or created or saved over a million jobs) both fail any analytical test.

A nit, but an important one to me. 40 percent of the stimulus was tax credits not tax cuts. A tax cuts requires you to be paying taxes to start with in order to receive it. A tax credit does not. A tax cut is often considered to be a rate change rather than a payment. It doesn't affect the macro point but it is important to keep in mind that the tax credit only would have a second order effect on job creation (spending multiplier) while a tax cut might change small business hiring plans (assuming of course that the business in question made so little money that it qualified).


What an Irreponsible Post, Stan

Stan, I wish you had taken the time to actually understand my stimulus argument before trashing it

Read this paper (which I have sent to you before), and then try again

http://www.heritage.org/Research/Economy/bg2354.cfm

Brian Riedl
The Heritage Foundation


You are so wrong Brian

You are making the claim in your paper that according to Keynesian theory "deficit" spending automatically stimulates the economy. That is a baseless claim. Nobody has ever made the claim.

If the deficit already exists at the beginning of the crisis it means little in terms of stimulating the economy beyond the current state. If President Obama were to have made a balanced budged his #1 priority as POTUS and took away that 1.2T deficit, the economy would have totally collapsed. It would have been devastating to the economy to just yank away that much money from states, businesses and people. The 1.2T 'stimulus' was already in place and the situation was dire. We needed more.

It's unbelievably dishonest of you to act as though anyone to prescribes to Keynesian thought believes that simply running a deficit means that everyone should be working.

I didn't get 2 paragraphs into your paper before I could see the partisan ramblings and heritage talking points shining through. At least try to be objective.


No Tim, you are wrong.

Tim, your post makes no sense from one paragraph to the next. Yot first paragraph states that Keynesianism doesn't say deficits=stimulus (wrong). Your second paragraph simply asserts (without any argument/evidence) that deficits definitely need to increase to stimulate the economy. Your third paragraph is ... well, I cannot even follow it.

Funny....and you call "my* writing partisan ramblings? That's all you can muster to refute my paper? Yikes.

Here's the reality: standard Keynesian theory defines fiscal "stimulus" as the increase in the budget deficit from one year to the next. It doesn't matter to Keynesians how the money is spent (his famous digging ditches example), nor does it matter whether the deficit increase is from a stimulus bill or from natural movements in spending and revenues (which is why the automatic triggering of higher unemployment insurance costs is called an "automatic stabilizer).

From 2008 through 2009, the deficit went from $450 billion to $1.4 trillion. This is nearly $1 trillion in "stimulus." Now if that were "new money" then GDP should have jumped by $1 trillion (actually $1.5 trillion if you count the multiplier). In other words, the economy should have overheated. It obviously didn't. Thus, standard Keynesian theory is wrong -- unless you believe the economy would have contracted an additional $1.5 trillion (11% of GDP) without the stimulus. Good luck arguing that.


Brian, I haven't yet read

Brian,

I haven't yet read either the your National Review piece or your Heritage paper, so forgive me if you make answers to the following clear, but...

You write:
From 2008 through 2009, the deficit went from $450 billion to $1.4 trillion. This is nearly $1 trillion in "stimulus." Now if that were "new money" then GDP should have jumped by $1 trillion (actually $1.5 trillion if you count the multiplier). In other words, the economy should have overheated. It obviously didn't. Thus, standard Keynesian theory is wrong -- unless you believe the economy would have contracted an additional $1.5 trillion (11% of GDP) without the stimulus. Good luck arguing that.

If we are talking about the net effect of the stimulus on the economy, we are speaking in ceteris paribus terms. So when you say "GDP should have jumped", aren't using an inappropriate reference point (GDP immediately prior to stimulus) rather than the appropriate reference point of the counterfactual (what GDP would have been without stimulus)?

And you then do use that appropriate (counterfactual) reference point, but I don't follow your reasoning: if GDP and employment were considerably below pre-recession levels (implying slack in the economy; excess capacity, output gap, etc.), why would the economy have to have "overheated" for it to be true that there was some net stimulative effect.

Notwithstanding the above, at the end of your paragraph you do focus on what seems to me the appropriate general question: What would GDP have been without the stimulus? And per my (limited) understanding of multipliers, I assume your premise is correct that if one is applying a 1.5 multiplier to stimulus (deficit) of $1 trillion, the implication is that GDP would be (or would have been) $1.5 trillion less without that stimulus, although I suppose that we need to be sure that the time periods are consistent (e.g., a 1.5 multiplier effect over 1 year has a different size impact on GDP in Year 1 than does a 1.5 multiplier effect over 2 years) -- I don't know if that's a problem with your particular figures or not (Do you think it is? Does the 1.5 you are using pertain to more than one year?). In any case, leaving aside the validity of someone's assertion of a particular multiplier for Year 1, all that would be needed for there to have been a stimulative effect of some magnitude would be a multiplier greater than zero and the counterfactual of a GDP that was even slightly lower than it actually was (not necessarily anywhere near 11% lower), right?


Good questions

Hi Brooks, good questions:

"And you then do use that appropriate (counterfactual) reference point, but I don't follow your reasoning: if GDP and employment were considerably below pre-recession levels (implying slack in the economy; excess capacity, output gap, etc.), why would the economy have to have "overheated" for it to be true that there was some net stimulative effect."

---- Here's my reasoning: The pre-stimulus GDP was probably about $300 billion below full-employment potential GDP. So $1 trillion in deficit spending ($1.5 trillion in multiplier-induced stimulus) should have closed the entire GDP gap and more. Yet the economy shrank further, leaving the output gap closer to $600-$800 billion.
In terms of counterfactuals, for the Keynesian theory to be correct, you'd have to believe the economy would have decreased an additional $1.5 trillion without all the new deficit spending. In other words, it requires believing that the economy would have decreased a staggering 13% of GDP without the new deficit stimulus, instead of the 3% of GDP drop we ended up with. I don't find that plausible

""Does the 1.5 you are using pertain to more than one year?). In any case, leaving aside the validity of someone's assertion of a particular multiplier for Year 1, all that would be needed for there to have been a stimulative effect of some magnitude would be a multiplier greater than zero and the counterfactual of a GDP that was even slightly lower than it actually was (not necessarily anywhere near 11% lower), right?""

------The multiplier of 1.5 – which I’m pulling from Keynesian economist Mark Zandi as well as the CBO – has appeared in their estimates of the present-year effect of the deficits, so I believe it’s a single-year figure.
Setting that aide,, you are correct, any multiplier greater than 0 is stimulative. Even a multiplier of 0.5 would mean a $1 trillion in stimulus raised GDP by $500 billion.
I believe (for reasons explained in my paper) that, yes, government spending circulates through the economy many times, creating a multiplier effect. But that’s money that (had it not been to the government) would have stayed in the private sector and been circulated with the same multiplier. So the act of government spending that dollar instead of the private sector makes no difference in the short-term GDP.

Best
Brian


No Brian...

Deficits have NOTHING to do with it, Brian. Deficits are merely an outcome as a result of added fiscal stimulus (or spending). You are conflating "deficits" with "spending". While, they do tend to be related they are simply not one in the same.

You are pretending that the increase in deficit was due to spending increases, which is false. Some of it was but the vast majority of it was due to the loss of tax receipts. Automatic stabilizers did kick in and did cost money, but they aren't designed to 'stimulate' they are designed to soften the blow of of the loss of income. An individual who could have purchased groceries and a big screen TV can now only afford the groceries. This is still a net loss to the economy.

Your concept of overheating the economy is also wrong:
1.) The deficit wasn't 'new spending' so you cannot treat it as such. Developing an argument based around that concept is terribly misleading.

2.)The concept of an overheating economy is based on a state of full employment - which is a state that we are nowhere close to. Our economy could safely take on double digit growth for quite some time without experiencing any 'over-heating'.

I am not going to pretend to be any kind of economic professor. I have a BA in Economics, but there is a lot I don't know. What I do know, is that you are either intentionally distorting Keynesian theory to fit your worldview or you have a deficit of knowledge when it comes to Keynesian thought.


Riedl and DeLong

I don't know enough to judge the quality of DeLong's critique, but I'm certain I don't care to read a piece which starts this way:

"Can We Please Shut National Review and the Heritage Foundation Down Now?"

"Stupidest and most intellectually dishonest thanktanks and magazines alive..."


Final response to Tim

Tim, I don't know how else to say this - you are wrong.

1) Keynesianism 101 says government spending injects money into the economy and taxes remove money from the economy. So the difference between those 2 figures (the budget deficit) is the net stimulus. The yearly movement of the deficit is thus the yearly change in stimulus. This is not a controversial point.
This means that, yes, all deficit increases are considered Keynesian stimulus. It doesn’t matter if the revenue decline is automatic or from a tax cut, Keynesian theory only cares that fewer money is being taxed from the economy. Similarly, it doesn't matter if the additional spending is from Unemployment rolls moving upward, or Congress passing a stimulus bill, as long as government is "injecting" that additional money into the economy. Thus, the movement in the deficit is the net stimulative effect. This is straight out of the General Theory, and straight out of Keynes' lectures.

2) You clearly missed my point on automatic stabilizers like unemployment insurance. Just because they don't solve a recession (obviously) does not mean that they are not considered "stimulus" under Keynesian theory. Anything that cushions the blow is considered "stimulus."

3) You also completely missed my point on overheating the economy. You say the stimulus could not overheat the economy since its in a recession. My point is that (by standard Keynesian theory) the $1 trillion surge in the budget deficit should have been big enough to end the recession, bring full employment, and overheat the economy. It didn’t.

4) Your related claim that "Our economy could safely take on double digit growth for quite some time without experiencing any 'over-heating' -- is bizarre.
Since the recession began, the economy has shrunk by 3%. The GDP is about 6% below what it would be at full-employment. So the idea the economy could start growing at a 10%+ annual rate for several years without overheating is simply silly. That claim has no basis in math, economics, or reality.

Of course, I’ll emphasize I don’t believe these Keynesian theories anyway. But your assertion that I misrepresented Keynesianism is flat out wrong.

Finally, you really need to drop the silly rhetoric describing any differing viewpoints as “partisan ramblings.” I know a lot of blog posters believe that insults and personal attacks make them look authoritative and tough. Instead, you just look intolerant and immature. People who have facts and logic on their side don’t need to resort to sophomoric insults.

Anyway, I’ve spent enough time on this mini-debate, so I’m checking out. Instead, I’ll wait to see if Stan actually gets around the reading the paper of mine that he slammed – which debunks the “idle savings” fallacy that he offered in his post.


Brian Riedl's big mistake is

That he doesn't seem to understand the concept of pareto efficiency. Or more precisely, what is a pareto efficient situation.

Let's assume that the pre-crisis was pareto efficient, ie all resources were deployed to maximum effect. The crisis causes demand to fall and production (supply) to fall with it.

Because production falls, resources are under-utilised. If I have a factory with 100 machines manned by 100 workers and sales fall by 20%, I fire 20 workers and mothball my 20 machines.

20 less workers are earning money, this impacts their spending, this impacts companies which in turn impacts their spending... etc etc. This is the negative multiplier.

Let's say I am the President and I personally give the factory owner an order equivalent to the 20% he previously lost.

Now, firstly I have 20 machines ready to use. It will cost me nothing (aside from maintenance and possibly electricity) to get them working. In other words, I can increase production by 20% for free! Furthermore, I can rehire my workers at a lower cost than previously. Why? Because the declining business spending (demand for jobs) and increased unemployed (supply of jobs) lowered the local labour cost.

So, I can run my profitable enterprise for less than before. My profits increase, workers get paid, yadda yadda yadda, positive multiplier effect.

Mr Riedl is confusing the situation outlined to the situation that would happen if I wanted to increase production in normal/good times. Then, I would have to hire my workers at a higher wage cost (because they are all employed) and I would also have to take out a loan to pay for my new machines.

Now of course, I have taken a great many simplifications and deviations from reality in my above example (chiefly that I am not the POTUS, that Congress does not exist and I do not own a factory) but the underlying point still remains:-

The first situation is a non-pareto efficient situation. I can literally get 20% more production at no fixed cost. I would be taking unallocated jobs and unallocated machines and using them. The second situation is a pareto efficient situation, I have to spend 20% more in fixed costs. That money is a change in the allocation of resources, I have to use allocated workers and buy more machines, which is an allocation of resources from someone else.

The first situation is positive-sum (contrary to what Mr Riedl asserts in his online paper), the second situation is zero-sum (as Mr Riedl correctly identifies in his online paper).

Mr Riedl is arguing that we are currently sitting in a pareto efficient situation, because any spending to increase GDP (by the government) would cause overheating. This is patently ridiculous, because to say that the economy would overheat means that actual GDP is exceeding potential GDP, ie we would be above trend GDP. If Mr Riedl knows one thing, I would hope it is that the economy has declined away from trend GDP.

My final point is about the ARRA and government spending. As Mr Riedl should know, for GDP to fall, so does the overall sum of Consumer Spending, Investment, Government Outlays and the balance of trade. If GDP were not to fall, the government would have to spend as much as the other three factors. What is the reality of the situation?

The reality is:
-GDP has fallen by almost 4% from peak.
-The ARRA total nominal value is $767bn or 5% of GDP.
-So far, 20% of that has been spent.
-Furthermore State/Local spending declined by 2.2% in FY2009 and is set to decline by 2.5% in FY2010.

So on my back of the napkin calculation, government spending has either been neutral or barely positive. So in other words, almost every contributing factor to GDP has fallen, and the one that hasn't has come close to offsetting the others.

Surely that's why we we're still in recession?


To Charles

"Mr Riedl is arguing that we are currently sitting in a pareto efficient situation, because any spending to increase GDP (by the government) would cause overheating."

Um....I said nothing of the sort. Actually my original point was government spending would have no stimulative effect whatsover. So obviously I don't believe it would cause overheating.

You may be thinking of my point that -- under (misguided)Keynesian theory -- $1 trillion in deficit spending would cause the economy to overheat (since it would provide $1.5 trillion in multiplier-induced stimulus, in an economy that is perhaps $800 billion below potential GDP).

That's obviously not the same thing as the saying the GDP is currently at optimal level and that ANY economic growth hike would cause overheating.

Also, again, you are defining the stimulus as ARRA only.
There's nothing in Keynesian theory that says "stimulus" is specifically defined as **only** those revenue/spending changes that occur in bills that Congress votes on (as opposed to automatic movements in revenues/spending), and only bills called "stimulus" bills in the media (as opposed to the multitude other tax/spending bills in 2009). So -- as I explained in detail above -- you have to count the entire $1 trillion increase in the deficit in 2009 as the "stimulus".

Best
Brian


To Charles

"Mr Riedl is arguing that we are currently sitting in a pareto efficient situation, because any spending to increase GDP (by the government) would cause overheating."

Um....I said nothing of the sort. Actually my original point was government spending would have no stimulative effect whatsover. So obviously I don't believe it would cause overheating.

You may be thinking of my point that -- under (misguided)Keynesian theory -- $1 trillion in deficit spending would cause the economy to overheat (since it would provide $1.5 trillion in multiplier-induced stimulus, in an economy that is perhaps $800 billion below potential GDP).

That's obviously not the same thing as the saying the GDP is currently at optimal level and that ANY economic growth hike would cause overheating.

Also, again, you are defining the stimulus as ARRA only.
There's nothing in Keynesian theory that says "stimulus" is specifically defined as **only** those revenue/spending changes that occur in bills that Congress votes on (as opposed to automatic movements in revenues/spending), and only bills called "stimulus" bills in the media (as opposed to the multitude other tax/spending bills in 2009). So -- as I explained in detail above -- you have to count the entire $1 trillion increase in the deficit in 2009 as the "stimulus".

Best
Brian


Brian, i can't believe you

Brian,
i can't believe you make the same stupid argument about crowding out that John Cochrane and Eugene Fama are making, that government borrowing necessarily crowds out private borrowing, even when the economy is depressed. Paul Krugman, (and i don't even like him very much most of the time) tears this one apart.
You don't seem to get than when the private sector is HOARDING MONEY, there isn't much investment going on in the first place. Also parking money in Treasury bills IS the functional equivalent (in the internet age) of hiding cash under the mattress (did you forget in September 2008, yields on T-bills went negative? And even now with the recovery under way, the fed funds rate is close to zero. How is that rational or an efficient market at work?) Investors are predictably irrational in times of trouble as a pack of lemmings. When bad times hit, they flee in terror to government bonds. which DON"T GET LENT OUT, unless stimulus occurs. Now you may have issues with the stimulus on what projects are worth investing in and what aren't That's a separate and and more worthwhile skeptical argument. But making the crowding out argument when the economy is depressed is inane. Incidently, if government borrowing is a zero-sum game with the private sector, than there's no reason why private sector borrowing isn't also.


Sigh

"You don't seem to get than when the private sector is HOARDING MONEY, there isn't much investment going on in the first place. Also parking money in Treasury bills IS the functional equivalent (in the internet age) of hiding cash under the mattress."

Sigh. Can you provide the locations of where these dollars are being hoarded? Mattresses? Safes? Where are they being shielded from all transactions? (rather than re-explain the "idle savings" fallacy again, just read this - http://tinyurl.com/yzccxax )

And if I buy a treasury bill from someone, I am obviously giving that person money to spend (and circulate through the economy)in return for the T-bill. If I buy a T-bill from the government, I am giving them money for spending.

How exactly is that hoarding dollars?

And, of course, I'm sure its too much to ask for you to actually the *read* the paper where I spell this all out (and address the specific issues you raise)
http://www.heritage.org/Research/Economy/bg2354.cfm

Brian


Question for Brian

Brian, I read your article, along with this one from Krugman: http://krugman.blogs.nytimes.com/2009/01/27/a-dark-age-of-macroeconomics.... The whole thing between the two of you seems like a chicken-or-egg argument to me.

Anyway, I'm no economist or anything, but your analysis of the expected effects from stimulus not appearing is compelling, definitely argues that the multiplier for that stimulus should be much lower than 1.5. But I'm struggling with other parts of your argument.

(1) Is cash in the mattress really the only way to hoard money? What about changes in the savings rate for individuals coupled with increased capital reserves at financial institutions? Isn't that money sitting on the sidelines? I suppose if that theory is right, we'd see large increases in the capital reserves at large financial institutions, something I have no idea how to check.

(2) Related to the above question RE: capital reserves, the story from the housing bubble seemed to boil down to the fact that financial institutions trusted their models too much and over-leveraged themselves. What does that loss of leverage do to the savings and investment rates?

Where I'm going with this is, is it possible that a huge chunk of the pre-stimulus GDP was just the result of the massive over-leveraging by the financial institutions, a bubble-GDP or a "paper" GDP. (I'm using "paper" n the same sense that an over-valued stock market produces "paper" wealth that evaporates when the stock comes back in line with more reasonable historical fundamentals.)

In other words, maybe real, non-paper GDP all along was much more than $300 billion below full-employment GDP, but the heavily leveraged financial institutions and their paper revenue and profits were masking it. I suppose this could also be verified by checking to see what percent of GDP was created by those firms during the bubble, and compare that percentage to historical norms for the sector.

Appreciate any response in advance.




Recent comments


Advertising


Order from Amazon


Copyright

Creative Commons LicenseThe content of CapitalGainsandGames.com is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License. Need permissions beyond the scope of this license? Please submit a request here.