Dean Baker reasons from a quantity change
I criticized Dean Baker for claiming that “all economists” believed you had to reduce the trade deficit to boost jobs. In the comment section he responded as follows:
Scott,
Glad to see that I seem to have got your juices flowing with this one. Let’s see if we can figure out the source of the misunderstanding.1) I assume the economy is below full employment in the standard Keynesian sense. Perhaps you don’t.
2) If it is below full employment then any increase in a component of demand (C,I,G, or [X-m]) will boost employment and output.
3) a main determinant of X-M is the price of goods and services in the United States relative to the price in other countries.
4) a lower valued dollar reduces the relative price of goods and services in the United States.
5) central banks can affect the relative price of currencies by buying or selling them.Now perhaps you can tell me which one(s) of these propositions you disagree with. FWIW, you will find versions of all of them in Greg Mankiw’s textbook, as I said in my post on his column.
My juices always flow when I see someone reason from a price or quantity change. And Baker shouldn’t feel bad, I’ve seen worse from Nobel Prize winners. But I’m afraid he’s not doing economics here, he’s doing accounting. Yes, GDP equals C + I + G + (X – M), but it’s also true that:
GDP = C + S + T
Does Baker want to claim that this proves that more saving leads to more GDP? How about higher taxes? As David Glasner likes to point out, accounting relationships tell us nothing about causation.
Intro textbooks often oversimplify macro at the beginning, saying things like more G increases GDP via a “multiplier,” even though the author knows that that relation assumes a fixed interest rate, not a 2% inflation target. And no crowding out. And no Ricardian equivalence. You are better off looking at Mankiw’s open economy macro chapters. In chapter 19 of the 3rd edition, Mankiw points out that bigger budget deficits lead to larger trade deficits, citing the famous example of the Reagan “twin deficits,” where deficit spending by Reagan boosted our trade deficit. Does Baker want to claim that a smaller trade deficit is expansionary if caused by fiscal austerity?
In his reply he sensibly shifts the argument to monetary policy. But Mankiw doesn’t even discuss monetary policy in the section on how shocks impact an open economy. That may be because (contrary to the claim of Baker) after monetary stimulus, the substitution effect (lower exchange rate) is often overwhelmed the by the income effects (a booming economy sucks in imports.) Yes, monetary stimulus boosts jobs in a depressed economy, but not because the trade deficit gets smaller. The most dramatic example of using monetary policy to devalue the dollar occurred in April 1933, and the trade deficit actually got “worse” over the next six months, as the resulting growth spurt sucked in imports. (Things slowed later with FDR’s high wage policy.) Needless to say not all textbooks get into this level of complexity at the intro level.
In August 1971, Nixon dramatically devalued the dollar 10%, ending Bretton Woods. The US trade deficit increased 4 fold in 1972 as compared to 1971, as very rapid GDP growth sucked in imports. In fairness, fiscal policy was also expansionary.
So on trade deficit changes caused by fiscal policy, Baker is flat out wrong. The theory on monetary induced changes is ambiguous, but there is lots of empirical evidence suggesting the income effect can often outweigh the substitution effect. And this isn’t just one of my crackpot theories, I’ve seen much more distinguished macroeconomists like Lars Svensson make the same point.
Of course Baker’s also wrong in claiming that any of this debate has any bearing on Mankiw’s discussion of the microeconomic efficiency gains from freer trade, which occur regardless of whether the economy is depressed.
Tags:
30. April 2015 at 06:20
He´s been at it for a long time.In 2013, reasoning from a price change:
“A lower-valued dollar will cause exports to soar and imports to plummet, creating millions of new manufacturing jobs. Millions more jobs would be created in other sectors due to the multiplier effect. This could well bring us back to full employment “” a goal we may not otherwise achieve until the next decade.”
https://thefaintofheart.wordpress.com/2013/10/10/dean-baker-gives-a-whole-new-meaning-to-creative-destruction/
30. April 2015 at 06:28
On a more general level: I don’t believe the issue is the benefits of trade. I would hope everyone realizes the while trade benefits an economy as a whole, there can be winners and losers and that it can be rational to decide the distribution of winners and losers can make a trade deal bad for the vast majority (perhaps not if the benefits were redistributed, but that’s not happening).
The current issues relate to the TPP. It is quite possible that some aspects of the TPP, such as intellectual property protectionism and dispute resolution provisions (aka loss of sovereignty), outweigh the trade aspects many economists are focusing on.
30. April 2015 at 06:41
‘In his reply he sensibly shifts the argument….’
Which is what you do when you know you’ve lost on the merits of the original argument.
30. April 2015 at 06:59
@Scott,
Does Dean admit that you’re an economist? Does he also admit that you disagree? If so, then he must retract his original statement.
30. April 2015 at 07:01
Reminds me of the many times someone would look at polyhedral magnetic well (aka polywell) fusion reactor concepts, and almost invariably they would immediately assume the electrons must collapse into a Debye sheath because the textbooks mainly deal with neutral plasma physics as that’s where nearly all the money has gone for about 40 years now. We’d have to explain that doesn’t happen when you’re operating a driven system at the quasineutral limit with a virtual anode. Then they’d stalk off muttering about ion currents flowing into the wall, which we never saw in experimentation. Details can be important!
30. April 2015 at 07:20
I am not sure Dean is looking through the proper lens. The sectoral balances view is what he should be looking at, and several factors are at play.
If you have a trade deficit then either the federal government or private sector must go into debt to offset that – has to net to zero – people must spend more than their income, if non residents spend less than their income in that currency.
The balances can be sustainable, without harm to employment.
Back before the GFC, we had a trade deficit of around 6% of GDP, and a budget deficit of only 1% that means the private sector had to go into debt to make up the difference – and they did. The debt levels were high relative to income, and without offset caused the economy to contract. This imbalance creates thin ice. Then say the fed tightens credit, or the fiscal position tightens and public sector’s income cannot support the private debt – then it can unravel things on the private sector side of the credit ledger. That credit contraction causes sales to fall and you will have an increased deficit and/or lower trade balance – and/or lower GDP.
So a trade deficit is not necessarily a bad thing, but has to be offset to and depending on what is happening in other areas. And it does not automatically cause unemployment.
30. April 2015 at 08:26
“Here’s What Wall Street Is Saying About The Sell-off in German Bunds”
http://www.bloomberg.com/news/articles/2015-04-30/here-s-what-wall-street-is-saying-about-that-drop-in-german-bunds
30. April 2015 at 08:39
C’mon,David, I think that’s a little hyperbolic. Economists both liberal or conservative-certainly including Scott-make comments about what ‘all economists agree on’ all the time, when you can find other economists that clearly disagree.
30. April 2015 at 09:40
First there are no good ways to manipulate the trade deficit directly. You can impose a tariff, for which your trading partners will institute a retaliatory tariff, and we get few goods trading, and higher prices, and effectively a supply shock. (See previous posts on correcting “oversupply” to fix a demand shortfall) Or, you can subsidize exporters, using good dollars to promote a less efficient production.
But more importantly, why would you want to? Since payments must be balanced, a trade flow must have a corresponding investment flow. If you restrict foreign trade, you will end up restricting foreign investment.
30. April 2015 at 10:43
Wikipedia: Twin deficits hypothesis – “What we can gather from this is the understanding of why an increased budget deficit goes up and down in tandem with the Trade Deficit. This is where we derive the appellation the Twin Deficits: if the US budget deficit goes up then either household savings must go up, the trade deficit must go up, or private investment will decrease.”
Sumner: “Does Baker want to claim that a smaller trade deficit is expansionary if caused by fiscal austerity?”
Seems Sumner is trying to put words in Baker’s mouth, as is typical sometimes, but indeed at full employment the above “Twin deficits hypothesis” suggests exactly what Sumner implies for Baker, in that private investment will increase with a smaller trade deficit and fiscal austerity, and since investment produces greater future GDP than private or public consumption, then it is indeed expansionary. But there’s all kinds of assumptions build into the above, including ones such as nothing the government does is truly additive to GDP in the long run, investment is expansionary to GDP, and the existence of “full potential output” (assumed in the Twin Deficits hypothesis).
30. April 2015 at 11:07
Scott,
you seem more interested in name calling than having a serious exchange. I laid out 5 propositions, which ones do you disagree with and why? That should be pretty simple.
btw, what I was referring to as something that all economists agree with is the accounting identity that GDP = C +I + G +(X-M). If you want to say you disagree with this identity, I stand corrected.
30. April 2015 at 11:16
Interesting gambit from Bernanke today;
http://www.brookings.edu/blogs/ben-bernanke/posts/2015/04/30-wsj-editorial-slow-growth-fed
‘The unemployment rate is a better indicator of cyclical conditions than the economic growth rate, and the relatively rapid decline in unemployment in recent years shows that the critical objective of putting people back to work is being met. Growth in output has been slow, despite solid job creation, because productivity gains have been slow””perhaps as the result of the financial crisis, which hammered new business formation and investment in research and development, perhaps for other reasons. But nobody claims that monetary policy can do much about productivity growth. Where it can be helpful is in supporting the return to full employment, and there the record has been reasonably good.’
Is that why he refuses to say why he doesn’t want to use NGDP as a target?
30. April 2015 at 11:23
Pot, Kettle, Black Award Committee, we have a winner;
‘Scott,
‘you seem more interested in name calling than having a serious exchange.’
That from the guy who started this with a post that contained;
‘It’s nice that Mankiw has apparently gotten out his bag of economist’s holy water….’
But, more importantly, just what name did Scott call you, Dean?
30. April 2015 at 11:43
better to learn history than accounting…
although the only thing we learn from history is…
30. April 2015 at 11:49
Ray, People were applying the twin deficits theory to Reagan when the unemployment rate was far higher than today. It does not assume full employment.
Dean, You said:
“you seem more interested in name calling”
Let’s see. You write a post insulting Mankiw, and then complain when I write a post in response that does NOT engage in name-calling, but does point out that you made a lot of silly EC101 level errors, like reasoning from an accounting identity.
I thought I answered your basic question, but since you didn’t see it that way let me spell it out as simply as I can:
1. I assume below full employment, if you wish.
2. No, that’s absurd, reasoning from an accounting identity.
3. No, it is determined by factors affecting national saving and national investment. The exchange rate is endogenous.
4. It depends on why the dollar fell, but your assumption is often correct in the short run.
5. I agree.
As far as Mankiw’s textbook, I addressed that in my post.
30. April 2015 at 11:52
DB,
Although that identity holds, its simply a definition of what GDP is in accounting terms, not a recipe for GDP growth.
For example:
GDP = C +I + G +(X-M)
If G were to increase and everything else held, then GDP would be higher. However lets say the government just kept ramping up spending to 100% of GDP. GDP may well (and most likely would be lower) look at the history of total state run economies. For example probably wouldn’t have higher GDP as consumption and investment would most likely crash.
I’m guessing personally you like the X-M part but again, nothing in this identity tells you how to increase GDP by fiddling with that part.
At any point in time of course that identity holds, by definition. But how best to grow GDP – well thats outside of anything this identity can tell you. (IANAE but they may call this exogenous)
30. April 2015 at 11:56
@Matt McOsker
-Where does NGDP targeting fit into this idea of yours?
@Dean Baker
-You still don’t get it, do you? Policies meant to affect only one part of an identity (say, net exports) have an affect on others (say, consumption). Economics consists of looking at unintended consequences.
30. April 2015 at 11:59
I usually agree with Baker, but I have to score this one for Scott. It must be gratifying for Scott to be able to call someone else out for a fallacious application of an accounting identity, since I seem to recall Krugman taking him to task for a similar blunder involving John Cochrane and the S=I identity a while back. It’s nice to see that he learned from that experience.
30. April 2015 at 12:23
E Harding – the fed would need to set policy that does any combination (1 or more) of the following – support public sector in maintaining deficits, enable (or not disrupt) private sector credit creation, devalue the dollar. Though it is not mutually exclusive to just these three items.
Using the sectoral balances I was just trying to illustrate that a trade deficit does not necessarily cause unemployment or slow growth if the slack is taken up by the private or public sectors.
30. April 2015 at 12:45
Interestingly, the US has made enormous strides in reducing the trade deficit–on the oil side.
Our oil deficit has fallen by $22 bn per month since early 2012. It is an astounding transformation for which the country’s shale industry deserves the credit.
However, since the beginning of 2013, improvements in the oil deficit has been offset virtually one-for-one with increases in deficits of other goods, with a non-oil deficit replacing an oil deficit.
There are three possible explanations.
First, it may be purely coincidental. Perhaps Americans decided to import more BMW’s just as the oil deficit lessened. Nevertheless, the timing is suspicious, leading us to believe that perhaps the total trade deficit may be the number to watch.
Thus, the US trade deficit may have a natural ‘comfort zone’, from which neither we nor our partners may feel compelled to deviate. This may be due to Americans’ confidence that they can handle increased debt associated with increased imports; or it may be due to a demand for US debt, such that foreigners are willing to depreciate their currencies to induce us to buy their goods and issue them debt.
Be that at is may, the US deficit appears quite resistant to improvement beyond a certain level, for reasons which may appear linked to exchange rates, but may reflect deeper relative demands for goods and debt.
30. April 2015 at 13:03
Steven – you could also look at it another way. The low price of oil will make shale development more expensive thus reducing domestic investment in shale exploration because it is too expensive. Will that reduce jobs despite a drop in the CAB? My guess is it could.
30. April 2015 at 14:17
@Dean Baker: Another answer to your (2), is monetary offset. Total aggregate demand is wherever the Fed wants it to be, regardless of what policymakers choose to do in specific industries. A government policy aimed at boosting exports, regardless what short-term pressure that exerts on NGDP, will not have an significant consequences to the economy, because of the Fed’s nominal target(s). Other components of GDP will simply adjust in order to reach the same total.
30. April 2015 at 14:18
Don’t expect Dean Baker to be persuaded by facts. I highlighted to him that his August 2002 paper that said housing was overvalued by 11% to 22% was wrong because house prices today are the same or slightly higher after inflation. Dean responded that this is only because interest rates are low. OMG, his defense is essentially that the “bubbble” didn’t pop because monetary policy was tighter than he expected.
30. April 2015 at 14:53
bill, low interest rates are not synonymous with low inflation.
30. April 2015 at 19:04
o. nate, Very funny, But Krugman lost that argument. I’ve never once made a unjustified causal argument from an accounting identity. If you can find an example I’ll put your name in highlights with an abject apology. Go to work!!
Krugman didn’t even read my post, he just relied on someone else who got it wrong.
I’m not as smart as Krugman, but please give me credit for not being a complete moron.
Bill, He seems to have overlooked the fact that rates are low (in part) because we had a Great Recession.
Now I’m beginning to see why the left revers Krugman so much. He seems to be about the only guy on their side that understands models.
1. May 2015 at 05:19
Okay I think I understand we are many trade advocates are coming from. Imagine, starting tomorrow, China India and Japan go crazy for American-made automobiles, Harleys and architects.
Suppose in next 4 months they buy $200 billion of new US automobiles motorcycles and architectural services. By the way today, architectural services can be done from home over the Internet.
Would not this flood of demand for US products and services boost US economic output and employment? Although it would not increase our living standards. We would do all that work for someone else’s benefit.
Interesting question.
1. May 2015 at 05:24
Ben, Yes, but suppose the trade deficit shrank because the US went into depression, and our consumers could no longer afford lots of imports. Would that help our economy?
1. May 2015 at 05:43
ben yes…i think the underlying feeling is the US (or any country) is like a family, we gotta bring more money in.
however, you shouldn’t reason about the US as a family, unless you reason about it as a 300 million person family, with lots of internal resources on the family owned land, the ability to print its own money, the ability of family members to provide other family members wants (services) etc etc
i like to disabuse of this notion by imagining the entire world (or imagine the US as the entire world if you like, theres no other countries) and say well, thered still be an economy, we’re not trading with mars.
what economic activity is, is everyone getting up everyday and creating wealth (making coffees, drilling for oil, cutting hair, waiting tables, designing robots, the robots making cars – all that stuff that people want) its not only about bringing home the bacon in the family analogy from somewhere else, which is the fallacious obsession with the X-M part.
1. May 2015 at 08:01
OT: http://www.moneyandbanking.com/commentary/2015/4/16/interview-with-adair-turner
“I am, however, very worried that we really don’t have an answer about what to do with the accumulated debt stock in the world. According to the 2014 Geneva Report and to the recent McKinsey Global Institute report, the total level of leverage in the world has simply continued to increase since 2008. We have not really achieved any deleveraging at all.” (Adair Turner, April 16, 2015, Senior Fellow, Institute for New Economic Thinking; former Chairman, U.K. Financial Services Authority; and former Chairman, Financial Stability Board Standing Committee on Supervisory and Regulatory Cooperation.)
By contrast, Sumner thinks central banks issuing more and more money cannot be bad. Who’s right? I am betting the smarter man is.
1. May 2015 at 10:48
Scott,
glad to see we’re making progress here.
I gave me third proposition as:
“a main determinant of X-M is the price of goods and services in the United States relative to the price in other countries.”
Your response is:
” No, it is determined by factors affecting national saving and national investment. The exchange rate is endogenous.”
okay, so do you want to give us the endogenous process that would reverse a sharp decline in foreign prices relative to U.S. prices, due to something like, a sharp rise in the value of the dollar relative to other currencies?
I’m also struck that you don’t believe that an exogenous increase in a component of demand in a below full employment economy leads to increased output (my proposition #2), but you’re welcome to believe whatever you like.
1. May 2015 at 17:00
“I’m also struck that you don’t believe that an exogenous increase in a component of demand in a below full employment economy leads to increased output (my proposition #2), but you’re welcome to believe whatever you like.”
-Dean, you just don’t get it, do you? If one sector of the economy is below full employment, that does not mean that all sectors of the economy are below full employment. So if government increases purchases of some rare and essential commodity with a vertical supply curve, that may well cause crowding out in the private sector, thus leading to lower total real GDP.
1. May 2015 at 17:03
“okay, so do you want to give us the endogenous process that would reverse a sharp decline in foreign prices relative to U.S. prices, due to something like, a sharp rise in the value of the dollar relative to other currencies?”
-Increases in capital outflows to those foreign countries with lower prices.
1. May 2015 at 20:54
If memory serves me-it’s tough to forget as so many posts are were written-David Glanser called you out for using an accounting identity to making an unjustified casual argument.
Or at least he might have-there were so many posts written it’s tough to remember who finally got the decision. LOL. I’m guessing that I’ll probably get a different answer from you than him.
2. May 2015 at 06:37
Dean, You said:
“okay, so do you want to give us the endogenous process that would reverse a sharp decline in foreign prices relative to U.S. prices, due to something like, a sharp rise in the value of the dollar relative to other currencies?”
I could spin lots of theories (like tight money, or rising productivity), but none would have any bearing on this discussion. Your point three is about X-M, not exchange rates. You seem to assume that a change in the exchange rate will cause a change in X-M, but that’s reasoning from a price change.
You said:
“I’m also struck that you don’t believe that an exogenous increase in a component of demand in a below full employment economy leads to increased output (my proposition #2), but you’re welcome to believe whatever you like.”
I don’t even know what that sentence means. What is an “exogenous increase” in X-M? If you mean an increase holding C+I+G constant, then you are right by definition, but that’s not what we were discussing.
By your logic “all economists” would agree that more G leads to more GDP, because G is a component of GDP. But it’s even worse, because I can sort of understand how someone could view G as an exogenous variable. But X-M?
Ray, You said:
“By contrast, Sumner thinks central banks issuing more and more money cannot be bad.”
Yes, I’m always praising the Zimbabwe central bank. Very funny.
Mike, You are mixed up. David was defining S & I differently. He didn’t think S=I is a identity, whereas the textbooks say it is. I was using the textbook definition. I think he’s backed off on his claim that I was making unjustified causal arguments based on an identity, but you’ll have to ask him. Based on my (and the textbook) definition, what I said was fine.
2. May 2015 at 06:38
Mike, And I’ve never made a “casual” argument. 🙂
2. May 2015 at 16:36
“By your logic “all economists” would agree that more G leads to more GDP, because G is a component of GDP. But it’s even worse, because I can sort of understand how someone could view G as an exogenous variable. But X-M?”
I’m at loss to understand why you can’t think of a change in X-M as exogenous. If China decides to lower the value of the dollar against the yuan, you don’t believe this leads to an increase in X-M for the U.S.?
As far as:
“”all economists” would agree that more G leads to more GDP, because G is a component of GDP.”
Yes, that is right — in an context where the economy is below full employment, which is my proposition. Do you want to tell me that you believe the economy can be below full employment in a Keynesian sense, but an increase in G doesn’t lead to an increase in GDP?
2. May 2015 at 17:26
Dean, just thought I’d say really cool to see you participating in the comments here, from a long time reader. Hope I’ll see more debates between you two in the future.
2. May 2015 at 22:14
Britonomist
There are lots of reasons why a rise in G might not boost GDP in an economy below capacity “in a Keynesian sense”. Most reasons have to do with expectations, a concept most Keynesians largely don’t accept has a place in their models.
Why not? I think it’s because Keynesians don’t understand or want to understand the positive but anarchic role of markets. Most Keynesians view the market as a blind engine of market failure. And Keynesians only see G as a conscious engine of success and never as a potentially disaster of unintended consequences and widely expected operational failure. Perhaps this is politics, I just think it is blindness.
Monetary offset is a subset of this bigger picture.
3. May 2015 at 05:54
Dean, You said:
“I’m at loss to understand why you can’t think of a change in X-M as exogenous. If China decides to lower the value of the dollar against the yuan, you don’t believe this leads to an increase in X-M for the U.S.?”
We are going around in circles here. I already explained to you that it makes all the difference in the world HOW the exchange rate change occurs. What does the government due to make it happen? Is it an easy money policy (as in Brazil) a high government saving policy (as in China)? You can’t just wave a magic wand and make your currency change. I’ve also suggested that when monetary policy is used to depreciate a currency the effect on X-M is ambiguous, due to the offsetting income and substitution effects. This is standard economics.
And even if X-M changes in a predictable way, it need not have any effect on GDP.
You said:
“Yes, that is right “” in an context where the economy is below full employment, which is my proposition. Do you want to tell me that you believe the economy can be below full employment in a Keynesian sense, but an increase in G doesn’t lead to an increase in GDP?”
Not only do I believe that, but it was pretty much the standard new Keynesian view up until 2007 (at least in terms of demand side effects, which is what I assume you are talking about, there are supply side effects on labor supply which could increase GDP, but those would probably lower economic welfare, by impoverishing the public.)
If the central bank is targeting inflation or NGDP, then an increase in G has no effect on inflation or NGDP, even if in a recession.
In 2013 the US did a massive amount of austerity, cutting the budget deficit by an astounding $500 billion compared to calendar 2012. Krugman said it would be a test of the austerity theory. A letter signed by 350 economists warned of possible recession. In contrast, I expected little or no effect due to monetary offset. Real GDP growth in 2013 was 3.13% compared to 1.67% in 2012. So much for Keynesianism.
Do you recall that in 2007 hardly any of the elite Keynesians still favored using activist fiscal policy in recessions? Do you recall Krugman mocking the Japanese for using fiscal stimulus in 1999, and asking why they didn’t just use good old fashioned monetary stimulus? And that was after he had written his famous 1998 liquidity trap paper.
Here’s Krugman in 1999, and my view today:
“What continues to amaze me is this: Japan’s current strategy of massive, unsustainable deficit spending in the hopes that this will somehow generate a self-sustained recovery is currently regarded as the orthodox, sensible thing to do – even though it can be justified only by exotic stories about multiple equilibria, the sort of thing you would imagine only a professor could believe. Meanwhile further steps on monetary policy – the sort of thing you would advocate if you believed in a more conventional, boring model, one in which the problem is simply a question of the savings-investment balance – are rejected as dangerously radical and unbecoming of a dignified economy.
Will somebody please explain this to me?”
Yes, and explain it to me too. Why did the economists profession lose its head after 2007?
3. May 2015 at 05:55
Here’s the Krugman link:
http://web.mit.edu/krugman/www/SCURVE.htm
3. May 2015 at 06:07
James, I think you meant to address that to Dean, but I’ll bite:
“Most reasons have to do with expectations, a concept most Keynesians largely don’t accept has a place in their models.”
Nice try, but New Keynesian economics is dominated by expectations based models.
I assume you’re referring to something like ricardian equivalence, but economists have good reasons to suspect this doesn’t always hold.
3. May 2015 at 13:07
The main Keynesian I read is Wren-Lewis and he fits my description perfectly.
3. May 2015 at 21:31
[…] Sumner – lots of my favorite posts of his are these shorter ones where someone has addressed a point he made, but is using faulty logic (or is reasoning from an identity). […]
4. May 2015 at 19:58
Frustrating to watch two brilliant guys (Scott and Dean) completely fail to engage with each other. If we could fix that, we could go so much faster.
5. May 2015 at 03:10
Britonomist
Can you recommend a good piece that brings the supply side into New Keynesianism?