Notwistanding Matt Yglesias, Noah Smith, and now Scott Sumner, I rather doubt that the belief that exports expand and imports contract income is due the the Y = C + I + G + (X-M) equation. It's just the businessman's POV. They can sell their wares to consumers or investors or the government or foreigners and imports of what they are selling reduce their market. The businessman usually sees the economy aa demand constrained; it's the economist that sees production as constrained by resources and technology.
"I rather doubt that the belief that exports expand and imports contract income is due the equation ... It's just the businessman's POV."
It's the view of a heck of a lot more people than that. Why do you think Trump is promising "Tariffs! Tariffs! Tariffs!" in every speech. Because it gets votes. If you are an average voter, you see imports coming in, those are things made abroad instead of here which costs us jobs here, while the money we are sending abroad to pay for them makes foreigners richer and leaves us poorer. Right?
But you are correct -- as far as the masses at Trump rallies are concerned, it's not the equation. How many have even ever seen it?
a) isn't it true that a reduction in imports has no effect on GDP only if you assume there's no substitution with domestically produced goods? A tarrif on European cars would reduce imports, but if consumers still spent the same amount on American cars, C wouldn't decrease, and GDP would go up, no?
b) S=I is true in the long run, but is it true in the short term? If people park more money in their bank, and their bank parks money at the Fed, how does that increase GDP?
a) You'd have to look at all the components of the equation. Obviously, GDP can go up at any given point in time. And obviously, it can do so at a time when M is falling. But the equation has no causal implications. If imports fall, do exports also fall? If not, does non-auto production also fall. My criticism is directed against those who believe the equation somehow shows a causal relationship. It doesn't.
b) No, it's an identity. It holds in both the short and long run. For any given money supply, if one person chooses to hold more, another person chooses to hold less.
On (b), maybe I'm being very slow here, but a consumer will either spend or save their money... If increasing savings is increasing I by the same amount, then no matter what a consumer does, they have no effect on GDP... What am I missing?
The term "saving" is a misleading concept when applied to the national income accounts. Suppose I receive $1000 and stick the money in a box. I might tell you that I've "saved" $1000. But in terms of the national income accounts, saving has not increased. The guy who gave me $1000 dissaved that same amount. In a simple model where there is no foreign trade or government: C + I = C + S
That means that S=I by definition. Saving is defined as the resources used to fund investment.
I am not an economist, but i have always thought that this is a central misunderstanding of Keynes. In the Great Depression, consumers were not just not spending, they were literally taking currency out of circulation and putting it under their mattresses . This problem was compounded by bank failures. (Milton Friedman appreciated this.) I suspect that Keynes' definition of saving was very different from what we now call savings. What do you think?
Understood, but few people these days stick money in boxes or under mattresses... so is it true that consumer choices don't really affect GDP all that much? Since either they spend their income, or they deposit it in the bank?
GDP is production. Thus to the extent that individual people have any impact on (real) GDP, it comes through their decision to work or to not work. Whether you spend or save your money probably has little or no effect on GDP, unless it somehow causes monetary policy to move off course. But even in that case, you don't know which way off course it would go.
Thanks for the response! Can't help feeling that the response to (a) isn't in good faith... Yes, I understand the equation doesn't show a casual relationship. But it's not exactly a leap of imagination to say that if people spend less on imports, they'll spend at least some of that money domestically. I feel that should be acknowledged.
Yes, but spending more domestically does not have any necessary implications for GDP. GDP measures production, not spending. Again, there are models where spending affects production, but when it does so it's not because these variables appear in the equation, it's because of factors such as sticky wages.
S ≠ I for the banks. S = I for the nonbanks. This is the source of the pervasive error that characterizes the Keynesian economics, the Gurley-Shaw thesis.
see: How Banks Create Money Out of NOTHING - Richard Werner (youtube.com)
Werner and Dr. Philip George haven't thought it all the way through. see: “The Riddle of Money Finally Solved”.
And as Dr. Philip George says; ““When interest rates go up, flows into savings and time deposits increase” ( the ratio of M1 to the sum of 12 months savings ).
Nicely explained, as to be expected. The equation might lead to populist rhetoric, but I think institutional failures lead to populist revolts.
Economists wildly under-estimating the costs to working class communities of badly done migration has a lot more to do with such. For instance, by zoning + migration driving up shelter costs (Kevin Erdmann has pointed out the shelter costs problem in the US). Or by breaking up local social capital. Or by suppressing wages through suppressing the Baumol effect. Or suppressing infrastructure production by increasing coordination problems. Or raising health costs by importing people who have been marrying their cousins for 1400 years. (Pakistanis are around 4% of births in the British population but generate around 30% of birth defects.)
Both Danish and Dutch research have shown that Middle Eastern Muslim migrants are a drain on the fisc at every age group. And, as we can see, they import the politics of the Middle East.
Robert Fogel wrote an entire book (‘Without Consent or Contract’) on how mass migration stressed the American republic along its fault line of slavery, and we can see migration stressing the US, UK and France along their metro/provincial fault-lines, and yet Bryan Caplan’s open borders idiocy is somehow intellectually respectable. Then again, Marxism is still intellectually respectable in universities, so …
This needs to be context specific. The cost benefit analysis of immigration depends on many national and even local level factors. How are immigrants attracted and selected? How well does the receiving area incentivize assimilation? What policies create or diminish any net fiscal benefit from immigrants?
Taking the US an an example, we don’t doe nearly enough to attract high-skilled, high potential immigrants, but do not do a good job of (de) selecting among asylum seekers. We do well however in assimilation and fiscal policies are good about creating net fiscal benefits.
A lot of commentary on migration is religious in the way a lot of commentary on socialism or communism is religious. So, just as socialism or communism never fail, people fail socialism, they fail communism, so migration never fails, people fail migration.
Just admitting that migration has costs, and these can be higher than the benefits, is sacrilegious for many folk, for whom politics has become their social salvation religion.
Migrants become a sacred category, meaning that even thinking about them in terms of trade-offs violates their sanctity: the sacred being that against which trade-offs must not be considered (except, maybe, against other sacred things). “Undocumented” maintains their sanctity, “illegal alien” profanes them.
Migrants tend to make where they come to more like the places they left.
The Latin Americanisation of the political economy of California—where you have a bloated, poorly functioning public sector and a rich, networked elite piling on regulation that squeezes out the middle class—has proceeded sufficiently, that folk are leaving it for the economic opportunities of the US.
I’m Australian. I look at how the US, UK and Western Europe “do” migration and am struck by how stupidly it is done. Canada used to do it sensibly, but PM Justin Blackface can screw anything up.
I certainly would not blame California's problems on Latin American immigrants. Relatively well run Orange County is 34% Hispanic, whereas San Francisco is only 15% Hispanic. Hispanics are not clamoring for woke policies like defund the police and stop enforcing shoplifting laws. I think you are just shooting from the hip, and are not well informed about the situation over here.
And Australia has a much higher rate of migration than the US and every single Oz city is safer than any US city. Strangely, how one does one’s policing turns out to matter. Who knew? But so does how one does one’s migration policy. https://www.lorenzofromoz.net/p/giving-thanks-for-arthur-calwell
Again, you are showing your ignorance of the situation in the US. Immigration here tends to reduce the crime rate. In addition, I have trouble seeing what this comment has to do with the post.
If you are Australian, you too are an immigrant unless you were one of the original inhabitants. Australia has grown a good bit in the past 200 years. Immigrants cannot have done too much damage.
My paternal ancestor arrived over 200 years ago, if you take time lines far enough back, everyone is an immigrant. Including various indigenous peoples. Who the immigrants are, how they arrive, under what circumstances, makes a difference. https://www.lorenzofromoz.net/p/giving-thanks-for-arthur-calwell
There is nothing wrong with the equation. It’s an accounting identity. What is wrong is using it to make statements about behavior using the identity. You need a more detailed model for that.
I thought this last equation was useful because then it (with, perhaps, accompanying circular flow diagram) can serve as a "frame" for students (or a dummy like me) to understand issues relating to the optimal level of investment. (Obviously it doesn't "explain" the level of investment).
"The important point is that [central banks] adjust policy to a position where they expect stable NGDP growth."
The Fed claims that it seeks PCE inflation at 2%, but is flexible about engineering more when deeded to facilitate adjustment to shocks and about how quickly to bring inflation back down to target. If successful, that probably rules out _wild_ swings in NGDP growth but I don't think they actually _target_ a stable growth rate.
A stable rate of NGDP growth is the logical way to address their dual mandate. I don't say that's exactly what they do, obviously, but it's a reasonable approximation of their goal.
Y = C+I+G+(X-M) is a statement about demand and not about output.
Then we take as tautology that supply must equal demand and income equal output.
Keynes gained his fame arguing that there are times when the market fails to clear and supply does not equal demand. But, the government (G) can manipulate the demand equation to bring the two sides of the equation into ballance.
He does not discuss whether there may be other levers in play, such as monetary policy, or to the degree that he does, he promotes fiscal policy as the his preferred mechanism.
Nor is there a suggestion that it is in the governments interest to try to influence the demand equation when the economy is at full employment or markets appear to be functioning.
I agree that the Keynes demand equation is not well taught to economics students, and it not properly understood by many professional economists.
One further beef -- the Keynesian model says that consumer demand is the most important component of demand. It is the largest component of demand, but it is not particularly dynamic. Investment is the dynamic component and deserves much more attention than it gets.
Great post! I remember very long ago encountering the equation and thinking it made no sense and of course the Prof in introductory econ never discussed it. Encountered it again in intermediate macro, money and banking, biz and gov course and along with it and a bit too much of IS-LM analysis decided that the engineering simple model approach to managing a complex economy was not for me so decided to do advanced studies in finance and not econ..
Nice to see that economists such as yourself are attempting to clarify the misunderstanding of this equation and a whole host of others in the field of economics.
I think from my experience a problem is too many people very naturally take "GDP = C + I + G + (X - M)" to be a model rather than a mere arithmetic description. E.g., I've heard many, many people say "Increasing G will increase GDP so more government please". But a model says more G comes from reducing C + I via taxes which impose deadweight loss, so smaller GDP may result. The standard disclaimer, "It's an accounting identity, don't reason from an accounting identity" is insufficient. Many undergrads and amateurs and Interwebers don't understand what an accounting identity is, or why one isn't a model. (Yes, MMTers, but many more.) I've known successful businesspeople who didn't understand how the accounting representations of their own businesses weren't models of the business. Take the rational ignorance of the masses seriously. It's best to Introduce this equation with KISS-level working examples of how using it as a model can go wrong. IMHO, FWIW.
Whether I responds to a drop in C depends on how well savings intermediation works in the economy. Wicksell talks about this in his lectures.
Take the extreme case of an economy where the MoE is the only savings vehicle available. A drop in C results in hoarding without providing cheaper funds to entrepreneurs.
I think the point you are making would be clearer if you simply suggested that a drop in C might be associated with a fall in NGDP. That's true. I don't believe "saving intermediation" is the issue, it just leads to confusion. Terms like "hoarding" are confusing, not clearly defined.
I was merely pointing out that the equation has no causal implications for a fall in C.
I was using hoarding as this the term Wicksell used in his lectures. I think defining hoarding as a sudden increase in the demand for money from sudden preference shifts is not ambigous and consistent with how past economists have used the term. Such hoarding results in lower NGDP.
But the intermediation angle remains important:
In a simple economy, where the only possibility to save funds is to accumulate money, an increase in savings is equal to an increase in money demand with the well known results. Thats the case Wicksell talks about.
In a modern economy, with many outlets for saving, increased demand for saving gets intermediated to entrepreneurs and a drop in C doesnt result in a drop in NGDP. ( as I increases). Only an increase in the demand for money (not savings) results in lower NGDP.
In case the intermediation function is impaired, this link breaks down...
Sure, increased demand for money can be deflationary (depending on how the Fed reacts.) But that has nothing to do with saving. Money hoarding isn't saving.
I'm actually not sure what you are disagreeing with.
Scott, not disagreeing (maybe on the importance of intermediation). Just trying to understand when a drop in C results in lower NGDP and when not, expanding on yout thoughts.
Saving is the deferal of current consumption for future consumption. Hoarding is a form of saving: By accumulating money, I store future purchasing power. So here I would disagree
C is 2/3rds of GDP, so obviously the two will be highly correlated. But the term "causation" is tricky. A drop in C doesn't just happen for no reason. There are bearish reasons (tight money) and bullish reasons (booming investment opportunities.) In WWII, C fell because G rose sharply. Again, that didn't cause a fall in GDP
Never reason from an identity. By which I mean do not give a behavioral interpretation to an identity. MMT does this a lot, saying that the fact that government deficits enter the national income identity and so do corporate profits means that government deficits raise corporate profits. This *can* happen, but it is not automatic the way reasoning from an identity implies.
Am I crazy for thinking that 99% of your disagreements with Keynesian economists are mostly just framing? It doesn’t seem like you disagree much on what’s actually happening. This is on the descriptive side. I do see some real prescriptive differences
Well, terms like "framing" and "actually happening" are kind of vague, so it's hard for me to answer your question. I use a different model---is that just different "framing"? I'm not sure.
Column idea; people who think we need to raise the minimum wage, but anything that raises wages for factory workers making $14 and hour is step one to Zimbabwe.
I don't disagree that the equation leads students - and frankly, generations of ordinary people (including economics, finance and business columnists and fund managers) - to misunderstandings about the economy. However, in partial defence of the equation, I would make the following points:
1. I typically think of AD on the left-hand side rather than GDP (or Q). My undergraduate macroeconomics courses focussed on what happens when AD is too low (or sometimes too high) to call forth sufficient production and income to yield full employment. And more generally, Keynesian economics is popularly known for its focus on 'demand deficiency' and conditions under which monetary policy could be ineffective; that is when money demand is relatively unresponsive to interest rates (ie a 'liquidity trap') and investment is relatively unresponsive to interest rates. It's the fault of later economists (and policy-makers) rather than Keynes for making that equation the centre of their workhorse model of business cycles.
2. By making G and the money supply exogenous, the equation facilitates the conception of fiscal and monetary policy as 'levers' that policy-makers can control. Interest rates and investment can change endogenously in response to exogenous variables like government spending or 'animal spirits', but the money supply remains fixed unless policy-makers decide otherwise. I can imagine that being a tremendously valuable aid for teaching students. Let's face it, almost everyone naturally thinks like the people of the concrete steppes.
3. Relatedly, the idea that policy-makers might respond to - and attempt to influence - market expectations came long after the AD=Y equation became the starting point for teaching macroeconomics. While students are taught and can understand the concept of 'automatic stabilisers' in fiscal policy, none would consider that those stabilisers would or could be set in such a way as to automatically equilibriate AD to full employment levels of output in the face of exogenous shocks. Similarly, while students accept the ability of central banks to respond to changes in demand or output, the idea that monetary policy could be devised in such a way as to keep nominal demand automatically on track in the face of shocks would until recent decades have been completely foreign. And this way of thinking persists, as you've no doubt found every time you try to persuade people that the Fed 'caused' the Great Recession! (And that's in spite of the fact that the same people would agree that a car driving off a twisty mountain road is probably the driver's fault.) Friedman and Schwartz had it much easier in 'A Monetary History' because they could show the money supply falling by one-third during the Great Depression. As well as naturally thinking in concrete steppes, ordinary people naturally refuse to accept the EMH, and accordingly refuse to believe that market expectations contain or reflect useful information that policy-makers should respond to agnostically; that is, even if those expectations cannot be logically and intuitively traced back to changes in 'fundamentals'.
So, in conclusion, I agree the equation is misleading, and textbook writers shouldn't use it - no one should. But neither Friedman and Schwartz nor their contemporaries or followers struck while the iron was hot and wrote a textbook setting out a fresh standalone model. Instead, New Keynesian economics developed and has generally done a ‘good-enough’ job to explain business cycles, and so the primacy of the demand-driven paradigm has persisted well beyond the refutation of its policy principles. Is it too late now more than four decades later to overthrow that paradigm? I would say so.
Savings= Investment, is an accounting identity, it can't tell you that "Thus if consumers spend less and save more, then C will decline but I will increase."
In fact changes in consumer spending have often caused recessions??
I think you are referring to the Keynesian concept of the paradox of thrift. This refers not to the case where consumers save more and cause a recession, rather it refers to a case where consumers try to save more, fail to do so, and cause a recession by contracting aggregate demand. But since I = S is an identity, if consumers actually do save more in the end, then investment goes up. That's what an identity means.
Yes I'm talking about the Paradox of thrift. For a given level of Investment, I, it is impossible for the country to collectively save more than that. If consumers reduce their consumption C, then that reduces that reduces national income 1 for 1, so that national savings remain unchanged. I wasn't aware that Keynes Paradox of thrift depended on a mote complicated argument then that, but I have only read a bit of Keynes.
Is the treatment of increased savings as a "leakage" that reduces the Keynesian multiplier (through a larger marginal propensity to save) part of "Vulgar Keynesianism?"
If people claim a long run effect, then it's vulgar Keynesianism. If short run, then it's Keynesianism. If short run only at the zero lower bound, then it's New Keynesianism.
Your view of macroeconomics seems to lead to the conclusion that monetary policy is omnipotent in steering nominal GDP. What about the effective lower bound on the policy interest rate?
Notwistanding Matt Yglesias, Noah Smith, and now Scott Sumner, I rather doubt that the belief that exports expand and imports contract income is due the the Y = C + I + G + (X-M) equation. It's just the businessman's POV. They can sell their wares to consumers or investors or the government or foreigners and imports of what they are selling reduce their market. The businessman usually sees the economy aa demand constrained; it's the economist that sees production as constrained by resources and technology.
"I rather doubt that the belief that exports expand and imports contract income is due the equation ... It's just the businessman's POV."
It's the view of a heck of a lot more people than that. Why do you think Trump is promising "Tariffs! Tariffs! Tariffs!" in every speech. Because it gets votes. If you are an average voter, you see imports coming in, those are things made abroad instead of here which costs us jobs here, while the money we are sending abroad to pay for them makes foreigners richer and leaves us poorer. Right?
But you are correct -- as far as the masses at Trump rallies are concerned, it's not the equation. How many have even ever seen it?
I have a couple of questions:
a) isn't it true that a reduction in imports has no effect on GDP only if you assume there's no substitution with domestically produced goods? A tarrif on European cars would reduce imports, but if consumers still spent the same amount on American cars, C wouldn't decrease, and GDP would go up, no?
b) S=I is true in the long run, but is it true in the short term? If people park more money in their bank, and their bank parks money at the Fed, how does that increase GDP?
a) You'd have to look at all the components of the equation. Obviously, GDP can go up at any given point in time. And obviously, it can do so at a time when M is falling. But the equation has no causal implications. If imports fall, do exports also fall? If not, does non-auto production also fall. My criticism is directed against those who believe the equation somehow shows a causal relationship. It doesn't.
b) No, it's an identity. It holds in both the short and long run. For any given money supply, if one person chooses to hold more, another person chooses to hold less.
On (b), maybe I'm being very slow here, but a consumer will either spend or save their money... If increasing savings is increasing I by the same amount, then no matter what a consumer does, they have no effect on GDP... What am I missing?
The term "saving" is a misleading concept when applied to the national income accounts. Suppose I receive $1000 and stick the money in a box. I might tell you that I've "saved" $1000. But in terms of the national income accounts, saving has not increased. The guy who gave me $1000 dissaved that same amount. In a simple model where there is no foreign trade or government: C + I = C + S
That means that S=I by definition. Saving is defined as the resources used to fund investment.
I am not an economist, but i have always thought that this is a central misunderstanding of Keynes. In the Great Depression, consumers were not just not spending, they were literally taking currency out of circulation and putting it under their mattresses . This problem was compounded by bank failures. (Milton Friedman appreciated this.) I suspect that Keynes' definition of saving was very different from what we now call savings. What do you think?
Understood, but few people these days stick money in boxes or under mattresses... so is it true that consumer choices don't really affect GDP all that much? Since either they spend their income, or they deposit it in the bank?
GDP is production. Thus to the extent that individual people have any impact on (real) GDP, it comes through their decision to work or to not work. Whether you spend or save your money probably has little or no effect on GDP, unless it somehow causes monetary policy to move off course. But even in that case, you don't know which way off course it would go.
Thanks for the response! Can't help feeling that the response to (a) isn't in good faith... Yes, I understand the equation doesn't show a casual relationship. But it's not exactly a leap of imagination to say that if people spend less on imports, they'll spend at least some of that money domestically. I feel that should be acknowledged.
Yes, but spending more domestically does not have any necessary implications for GDP. GDP measures production, not spending. Again, there are models where spending affects production, but when it does so it's not because these variables appear in the equation, it's because of factors such as sticky wages.
S ≠ I for the banks. S = I for the nonbanks. This is the source of the pervasive error that characterizes the Keynesian economics, the Gurley-Shaw thesis.
see: How Banks Create Money Out of NOTHING - Richard Werner (youtube.com)
Werner and Dr. Philip George haven't thought it all the way through. see: “The Riddle of Money Finally Solved”.
And as Dr. Philip George says; ““When interest rates go up, flows into savings and time deposits increase” ( the ratio of M1 to the sum of 12 months savings ).
This of course explains Japan's "lost decade".
Nicely explained, as to be expected. The equation might lead to populist rhetoric, but I think institutional failures lead to populist revolts.
Economists wildly under-estimating the costs to working class communities of badly done migration has a lot more to do with such. For instance, by zoning + migration driving up shelter costs (Kevin Erdmann has pointed out the shelter costs problem in the US). Or by breaking up local social capital. Or by suppressing wages through suppressing the Baumol effect. Or suppressing infrastructure production by increasing coordination problems. Or raising health costs by importing people who have been marrying their cousins for 1400 years. (Pakistanis are around 4% of births in the British population but generate around 30% of birth defects.)
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4567984/
Both Danish and Dutch research have shown that Middle Eastern Muslim migrants are a drain on the fisc at every age group. And, as we can see, they import the politics of the Middle East.
Robert Fogel wrote an entire book (‘Without Consent or Contract’) on how mass migration stressed the American republic along its fault line of slavery, and we can see migration stressing the US, UK and France along their metro/provincial fault-lines, and yet Bryan Caplan’s open borders idiocy is somehow intellectually respectable. Then again, Marxism is still intellectually respectable in universities, so …
This needs to be context specific. The cost benefit analysis of immigration depends on many national and even local level factors. How are immigrants attracted and selected? How well does the receiving area incentivize assimilation? What policies create or diminish any net fiscal benefit from immigrants?
Taking the US an an example, we don’t doe nearly enough to attract high-skilled, high potential immigrants, but do not do a good job of (de) selecting among asylum seekers. We do well however in assimilation and fiscal policies are good about creating net fiscal benefits.
I hear that things are very different in UK.
A lot of commentary on migration is religious in the way a lot of commentary on socialism or communism is religious. So, just as socialism or communism never fail, people fail socialism, they fail communism, so migration never fails, people fail migration.
Just admitting that migration has costs, and these can be higher than the benefits, is sacrilegious for many folk, for whom politics has become their social salvation religion.
Migrants become a sacred category, meaning that even thinking about them in terms of trade-offs violates their sanctity: the sacred being that against which trade-offs must not be considered (except, maybe, against other sacred things). “Undocumented” maintains their sanctity, “illegal alien” profanes them.
Migrants tend to make where they come to more like the places they left.
https://www.sup.org/books/economics-and-finance/culture-transplant
So you have Muslim localities in Europe where public spaces are male dominated to the extent of excluding women.
https://www.harpercollins.com/products/prey-ayaan-hirsi-ali?variant=32126595203106
The Latin Americanisation of the political economy of California—where you have a bloated, poorly functioning public sector and a rich, networked elite piling on regulation that squeezes out the middle class—has proceeded sufficiently, that folk are leaving it for the economic opportunities of the US.
I’m Australian. I look at how the US, UK and Western Europe “do” migration and am struck by how stupidly it is done. Canada used to do it sensibly, but PM Justin Blackface can screw anything up.
I certainly would not blame California's problems on Latin American immigrants. Relatively well run Orange County is 34% Hispanic, whereas San Francisco is only 15% Hispanic. Hispanics are not clamoring for woke policies like defund the police and stop enforcing shoplifting laws. I think you are just shooting from the hip, and are not well informed about the situation over here.
As noted above, San Diego is the country's safest large city. It has its share of recent immigrants.
And Australia has a much higher rate of migration than the US and every single Oz city is safer than any US city. Strangely, how one does one’s policing turns out to matter. Who knew? But so does how one does one’s migration policy. https://www.lorenzofromoz.net/p/giving-thanks-for-arthur-calwell
Again, you are showing your ignorance of the situation in the US. Immigration here tends to reduce the crime rate. In addition, I have trouble seeing what this comment has to do with the post.
If you are Australian, you too are an immigrant unless you were one of the original inhabitants. Australia has grown a good bit in the past 200 years. Immigrants cannot have done too much damage.
My paternal ancestor arrived over 200 years ago, if you take time lines far enough back, everyone is an immigrant. Including various indigenous peoples. Who the immigrants are, how they arrive, under what circumstances, makes a difference. https://www.lorenzofromoz.net/p/giving-thanks-for-arthur-calwell
There is nothing wrong with the equation. It’s an accounting identity. What is wrong is using it to make statements about behavior using the identity. You need a more detailed model for that.
Yes, it's technically correct. But I would argue that this equation does not have a single useful purpose in economic textbooks.
GDP = C + I + G + (X - M)
GDP = C + S + T
Therefore:
I = S + (M - X) - (G - T)
I had always thought this was useful - it isn't?
I thought this last equation was useful because then it (with, perhaps, accompanying circular flow diagram) can serve as a "frame" for students (or a dummy like me) to understand issues relating to the optimal level of investment. (Obviously it doesn't "explain" the level of investment).
I = S + (M - X) - (G - T)
Yes, it may be useful to think of investment as reflecting private saving plus government saving plus foreign saving.
"The important point is that [central banks] adjust policy to a position where they expect stable NGDP growth."
The Fed claims that it seeks PCE inflation at 2%, but is flexible about engineering more when deeded to facilitate adjustment to shocks and about how quickly to bring inflation back down to target. If successful, that probably rules out _wild_ swings in NGDP growth but I don't think they actually _target_ a stable growth rate.
A stable rate of NGDP growth is the logical way to address their dual mandate. I don't say that's exactly what they do, obviously, but it's a reasonable approximation of their goal.
see N-gDp growth rate:
https://fred.stlouisfed.org/series/GDP
The acceleration in the upward sloping curve clearly demonstrates an excessively easy money policy.
I agree it’s not bad, but my question is, why is it better than FAIT?
NGDPLT is one type of FAIT. I think it's better than the actual type of FAIT that the Fed implemented, which resulted in high inflation.
But if you wish to call NGDPLT flexible inflation targeting, that's fine with me.
Y = C+I+G+(X-M) is a statement about demand and not about output.
Then we take as tautology that supply must equal demand and income equal output.
Keynes gained his fame arguing that there are times when the market fails to clear and supply does not equal demand. But, the government (G) can manipulate the demand equation to bring the two sides of the equation into ballance.
He does not discuss whether there may be other levers in play, such as monetary policy, or to the degree that he does, he promotes fiscal policy as the his preferred mechanism.
Nor is there a suggestion that it is in the governments interest to try to influence the demand equation when the economy is at full employment or markets appear to be functioning.
I agree that the Keynes demand equation is not well taught to economics students, and it not properly understood by many professional economists.
One further beef -- the Keynesian model says that consumer demand is the most important component of demand. It is the largest component of demand, but it is not particularly dynamic. Investment is the dynamic component and deserves much more attention than it gets.
Great post! I remember very long ago encountering the equation and thinking it made no sense and of course the Prof in introductory econ never discussed it. Encountered it again in intermediate macro, money and banking, biz and gov course and along with it and a bit too much of IS-LM analysis decided that the engineering simple model approach to managing a complex economy was not for me so decided to do advanced studies in finance and not econ..
Nice to see that economists such as yourself are attempting to clarify the misunderstanding of this equation and a whole host of others in the field of economics.
I think from my experience a problem is too many people very naturally take "GDP = C + I + G + (X - M)" to be a model rather than a mere arithmetic description. E.g., I've heard many, many people say "Increasing G will increase GDP so more government please". But a model says more G comes from reducing C + I via taxes which impose deadweight loss, so smaller GDP may result. The standard disclaimer, "It's an accounting identity, don't reason from an accounting identity" is insufficient. Many undergrads and amateurs and Interwebers don't understand what an accounting identity is, or why one isn't a model. (Yes, MMTers, but many more.) I've known successful businesspeople who didn't understand how the accounting representations of their own businesses weren't models of the business. Take the rational ignorance of the masses seriously. It's best to Introduce this equation with KISS-level working examples of how using it as a model can go wrong. IMHO, FWIW.
Scott,
Whether I responds to a drop in C depends on how well savings intermediation works in the economy. Wicksell talks about this in his lectures.
Take the extreme case of an economy where the MoE is the only savings vehicle available. A drop in C results in hoarding without providing cheaper funds to entrepreneurs.
I think the point you are making would be clearer if you simply suggested that a drop in C might be associated with a fall in NGDP. That's true. I don't believe "saving intermediation" is the issue, it just leads to confusion. Terms like "hoarding" are confusing, not clearly defined.
I was merely pointing out that the equation has no causal implications for a fall in C.
Scott,
I was using hoarding as this the term Wicksell used in his lectures. I think defining hoarding as a sudden increase in the demand for money from sudden preference shifts is not ambigous and consistent with how past economists have used the term. Such hoarding results in lower NGDP.
But the intermediation angle remains important:
In a simple economy, where the only possibility to save funds is to accumulate money, an increase in savings is equal to an increase in money demand with the well known results. Thats the case Wicksell talks about.
In a modern economy, with many outlets for saving, increased demand for saving gets intermediated to entrepreneurs and a drop in C doesnt result in a drop in NGDP. ( as I increases). Only an increase in the demand for money (not savings) results in lower NGDP.
In case the intermediation function is impaired, this link breaks down...
Sure, increased demand for money can be deflationary (depending on how the Fed reacts.) But that has nothing to do with saving. Money hoarding isn't saving.
I'm actually not sure what you are disagreeing with.
Scott, not disagreeing (maybe on the importance of intermediation). Just trying to understand when a drop in C results in lower NGDP and when not, expanding on yout thoughts.
Saving is the deferal of current consumption for future consumption. Hoarding is a form of saving: By accumulating money, I store future purchasing power. So here I would disagree
C is 2/3rds of GDP, so obviously the two will be highly correlated. But the term "causation" is tricky. A drop in C doesn't just happen for no reason. There are bearish reasons (tight money) and bullish reasons (booming investment opportunities.) In WWII, C fell because G rose sharply. Again, that didn't cause a fall in GDP
Never reason from an identity. By which I mean do not give a behavioral interpretation to an identity. MMT does this a lot, saying that the fact that government deficits enter the national income identity and so do corporate profits means that government deficits raise corporate profits. This *can* happen, but it is not automatic the way reasoning from an identity implies.
Yes, MMT is kind of a disaster with respect to reasoning from identities.
Am I crazy for thinking that 99% of your disagreements with Keynesian economists are mostly just framing? It doesn’t seem like you disagree much on what’s actually happening. This is on the descriptive side. I do see some real prescriptive differences
Well, terms like "framing" and "actually happening" are kind of vague, so it's hard for me to answer your question. I use a different model---is that just different "framing"? I'm not sure.
Column idea; people who think we need to raise the minimum wage, but anything that raises wages for factory workers making $14 and hour is step one to Zimbabwe.
I don't disagree that the equation leads students - and frankly, generations of ordinary people (including economics, finance and business columnists and fund managers) - to misunderstandings about the economy. However, in partial defence of the equation, I would make the following points:
1. I typically think of AD on the left-hand side rather than GDP (or Q). My undergraduate macroeconomics courses focussed on what happens when AD is too low (or sometimes too high) to call forth sufficient production and income to yield full employment. And more generally, Keynesian economics is popularly known for its focus on 'demand deficiency' and conditions under which monetary policy could be ineffective; that is when money demand is relatively unresponsive to interest rates (ie a 'liquidity trap') and investment is relatively unresponsive to interest rates. It's the fault of later economists (and policy-makers) rather than Keynes for making that equation the centre of their workhorse model of business cycles.
2. By making G and the money supply exogenous, the equation facilitates the conception of fiscal and monetary policy as 'levers' that policy-makers can control. Interest rates and investment can change endogenously in response to exogenous variables like government spending or 'animal spirits', but the money supply remains fixed unless policy-makers decide otherwise. I can imagine that being a tremendously valuable aid for teaching students. Let's face it, almost everyone naturally thinks like the people of the concrete steppes.
3. Relatedly, the idea that policy-makers might respond to - and attempt to influence - market expectations came long after the AD=Y equation became the starting point for teaching macroeconomics. While students are taught and can understand the concept of 'automatic stabilisers' in fiscal policy, none would consider that those stabilisers would or could be set in such a way as to automatically equilibriate AD to full employment levels of output in the face of exogenous shocks. Similarly, while students accept the ability of central banks to respond to changes in demand or output, the idea that monetary policy could be devised in such a way as to keep nominal demand automatically on track in the face of shocks would until recent decades have been completely foreign. And this way of thinking persists, as you've no doubt found every time you try to persuade people that the Fed 'caused' the Great Recession! (And that's in spite of the fact that the same people would agree that a car driving off a twisty mountain road is probably the driver's fault.) Friedman and Schwartz had it much easier in 'A Monetary History' because they could show the money supply falling by one-third during the Great Depression. As well as naturally thinking in concrete steppes, ordinary people naturally refuse to accept the EMH, and accordingly refuse to believe that market expectations contain or reflect useful information that policy-makers should respond to agnostically; that is, even if those expectations cannot be logically and intuitively traced back to changes in 'fundamentals'.
So, in conclusion, I agree the equation is misleading, and textbook writers shouldn't use it - no one should. But neither Friedman and Schwartz nor their contemporaries or followers struck while the iron was hot and wrote a textbook setting out a fresh standalone model. Instead, New Keynesian economics developed and has generally done a ‘good-enough’ job to explain business cycles, and so the primacy of the demand-driven paradigm has persisted well beyond the refutation of its policy principles. Is it too late now more than four decades later to overthrow that paradigm? I would say so.
I'm fine with a "demand driven" approach to macro, I'd just prefer they not use this equation. Might I suggest MV = PY?
* Correction - a liquidity trap of course means interest rates are relatively unresponsive to the money supply.
Hang on,
Savings= Investment, is an accounting identity, it can't tell you that "Thus if consumers spend less and save more, then C will decline but I will increase."
In fact changes in consumer spending have often caused recessions??
I think you are referring to the Keynesian concept of the paradox of thrift. This refers not to the case where consumers save more and cause a recession, rather it refers to a case where consumers try to save more, fail to do so, and cause a recession by contracting aggregate demand. But since I = S is an identity, if consumers actually do save more in the end, then investment goes up. That's what an identity means.
Yes I'm talking about the Paradox of thrift. For a given level of Investment, I, it is impossible for the country to collectively save more than that. If consumers reduce their consumption C, then that reduces that reduces national income 1 for 1, so that national savings remain unchanged. I wasn't aware that Keynes Paradox of thrift depended on a mote complicated argument then that, but I have only read a bit of Keynes.
I agree that if consumption falls and saving is unchanged then GDP falls. I was speaking to the case where saving increases.
Is the treatment of increased savings as a "leakage" that reduces the Keynesian multiplier (through a larger marginal propensity to save) part of "Vulgar Keynesianism?"
If people claim a long run effect, then it's vulgar Keynesianism. If short run, then it's Keynesianism. If short run only at the zero lower bound, then it's New Keynesianism.
Thanks for the explanation.
Your view of macroeconomics seems to lead to the conclusion that monetary policy is omnipotent in steering nominal GDP. What about the effective lower bound on the policy interest rate?
That's not a constraint, as monetary policy is fundamentally about the supply and demand for base money, not interest rates.