3 ms·

Aggregate demand is a money quantity, so it is perfectly well-defined. Keynes never tries to add 1 slice of pizza to 1 computer, and in fact goes into great det

Aggregate demand is a money quantity, so it is perfectly well-defined. Keynes never tries to add 1 slice of pizza to 1 computer, and in fact goes into great detail about units and definitions in book 2 ("definitions and ideas") of The General Theory. An uncharitable person would say you've probably never even read it.

yummyfajitas 16y ago
I did not claim Keynes ignored units. I claimed he reduced the economy to a few aggregate quantities and claimed they are sufficient to make predictions. Statistical physics does much the same thing - they reduce a complex newtonian system of 10^23 particles to a few aggregate quantities: temperature, pressure, etc, and makes predictions based on the aggregates. Keynes tries to do the same thing, reducing a bunch of sectorial demand curves into a single aggregate demand curve. Austrians/Chicagoans believe this reduction is unjustified. I freely admit I've only read a small portion of General Theory, though I've read a bit more by newer Keynesians (Krugman and Thoma are two names that spring to mind).
- jhpriestley 16y ago
  Okay, I assumed you were talking about the 'aggregation problem' (the aggregation of unlike quantities, rather a common pitfall in economics). If you agree that aggregate demand is a well-defined quantity, then I'm not sure what you're objecting to. You can disagree with a theory based on this curve, but the curve itself requires no justification beyond a proof that it is well-defined.
  - yummyfajitas 16y ago
    An aggregate theory of any sort has two separate pieces. The first is the definition of the aggregate. The second is a a reduction of the more complicated system to a simplified one based on the aggregate, i.e. showing that theories based on the aggregate quantity are sufficient to predict the world. I'll make the analogy to statistical physics again. You can always define the average temperature and pressure in a vessel. One can come up with simple calculus laws, such as PV=nT and make predictions(remember high school chem?). Sometimes that's an oversimplification, since temperature and pressure vary from place to place within the vessel. Sometimes you can't even use simple gas laws and need to revert to newtonian or quantum mechanics. The belief of Austrians (and myself) is that for the most part, the simple calculus of aggregates is an oversimplification and hides important details necessary to explain the world. One example: you can make a two-sector model of the economy, say [housing, everything else]. In this case, demand is a vector, say [housing demand, other demand]. The demand shock is in housing, and that's where a lot of the unemployment lives (there is some unemployment in "everything else" due to knock-on effects). It's fairly clear to see that stimulus will have different effects depending on whether it is a housing stimulus or an "everything else" stimulus, and that stimulus can mostly only help with the knock-on effects. A theory based solely on aggregate demand is completely incapable of recognizing this, since it doesn't even have the variables to describe it. The AD theory simply studies AD=housing demand + everything else demand, and tries to go from there. If, for example, housing demand = alpha x AD, and everything else demand = beta x AD, an AD-based theory would be very useful since it would just hide unnecessary complexity. If that isn't the case, then an AD-based theory hides necessary complexity and gives the wrong answers.
    - jhpriestley 16y ago
      It's fairly clear to see that stimulus will have different effects depending on whether it is a housing stimulus or an "everything else" stimulus, and that stimulus can mostly only help with the knock-on effects. This is exactly what is not clear. If Keynes is right, then any stimulus which boosts aggregate demand will have the same effect on unemployment, whether it acts through the housing market or by building Furbies. It is not that Keynes is unable to recognize the distinction; he positively asserts that the distinction does not matter when in a situation of deficient aggregate demand. A lot of the unemployment is in housing and related industries (~2 million jobs lost in construction), but a lot of it is also not in housing and related industries (~8 million jobs lost total). Every industry save health and education has lost jobs since 2008. The expected effect from a large loss of jobs in one sector, absent the resulting loss in AD, would be an increase of employment in other sectors, since the employees freed from the bad sector would lower the price of labor in other sectors. This is simple supply and demand. The fact that the opposite impact is seen every single time strongly suggests that the adjustment is blocked, per Keynes theory.
      - yummyfajitas 16y ago
        That's exactly what I was saying. Keynes assumes he can reduce the world to Calc 101. Austrians/Chicagoans/etc don't. As for why I think it's clear this reduction is not always valid, a simple hypothetical: suppose the stimulus targeted doctor-delivered medicine. There is no unemployment among doctors, so there is no pool of underutilized doctors to employ. It also takes about 8 years to train a doctor, so any effect that stimulus has on inducing construction workers (or others) to become doctors can only occur 8 years later. Also, sticky wages don't immediately lead to the validity of AD or keynesianism. For example, suppose construction workers earned $30/hour before the bust and assume Keynesian wage stickiness. They will not take un-stimulated jobs at Walmart at $12/hour or even stimulated jobs at $24/hour and they will be unqualified for Sous Chef jobs at $30/hour or even a stimulated $35/hour. Unless there is some specific job out there that pays at least $30/hour and construction workers are qualified for it, they will simply remain unemployed. Thus, under some circumstances, Keynes own assumptions make his Calc 101 reduction invalid. Lastly, there is no reason a housing bubble bursting would cause job losses only in construction. It would also harm realtors, mortgage brokers, bankers, etc. It would indirectly harm people who sell things to realtors, mortgage brokers and bankers. Thus, in the recalculation picture, we'd expect the biggest job losses in construction, real estate and mortgages, and correspondingly smaller job losses as you move outward through the economy from this epicenter.

yummyfajitas 16y ago

I did not claim Keynes ignored units. I claimed he reduced the economy to a few aggregate quantities and claimed they are sufficient to make predictions. Statistical physics does much the same thing - they reduce a complex newtonian system of 10^23 particles to a few aggregate quantities: temperature, pressure, etc, and makes predictions based on the aggregates. Keynes tries to do the same thing, reducing a bunch of sectorial demand curves into a single aggregate demand curve. Austrians/Chicagoans believe this reduction is unjustified. I freely admit I've only read a small portion of General Theory, though I've read a bit more by newer Keynesians (Krugman and Thoma are two names that spring to mind).

jhpriestley 16y ago
Okay, I assumed you were talking about the 'aggregation problem' (the aggregation of unlike quantities, rather a common pitfall in economics). If you agree that aggregate demand is a well-defined quantity, then I'm not sure what you're objecting to. You can disagree with a theory based on this curve, but the curve itself requires no justification beyond a proof that it is well-defined.
- yummyfajitas 16y ago
  An aggregate theory of any sort has two separate pieces. The first is the definition of the aggregate. The second is a a reduction of the more complicated system to a simplified one based on the aggregate, i.e. showing that theories based on the aggregate quantity are sufficient to predict the world. I'll make the analogy to statistical physics again. You can always define the average temperature and pressure in a vessel. One can come up with simple calculus laws, such as PV=nT and make predictions(remember high school chem?). Sometimes that's an oversimplification, since temperature and pressure vary from place to place within the vessel. Sometimes you can't even use simple gas laws and need to revert to newtonian or quantum mechanics. The belief of Austrians (and myself) is that for the most part, the simple calculus of aggregates is an oversimplification and hides important details necessary to explain the world. One example: you can make a two-sector model of the economy, say [housing, everything else]. In this case, demand is a vector, say [housing demand, other demand]. The demand shock is in housing, and that's where a lot of the unemployment lives (there is some unemployment in "everything else" due to knock-on effects). It's fairly clear to see that stimulus will have different effects depending on whether it is a housing stimulus or an "everything else" stimulus, and that stimulus can mostly only help with the knock-on effects. A theory based solely on aggregate demand is completely incapable of recognizing this, since it doesn't even have the variables to describe it. The AD theory simply studies AD=housing demand + everything else demand, and tries to go from there. If, for example, housing demand = alpha x AD, and everything else demand = beta x AD, an AD-based theory would be very useful since it would just hide unnecessary complexity. If that isn't the case, then an AD-based theory hides necessary complexity and gives the wrong answers.
  - jhpriestley 16y ago
    It's fairly clear to see that stimulus will have different effects depending on whether it is a housing stimulus or an "everything else" stimulus, and that stimulus can mostly only help with the knock-on effects. This is exactly what is not clear. If Keynes is right, then any stimulus which boosts aggregate demand will have the same effect on unemployment, whether it acts through the housing market or by building Furbies. It is not that Keynes is unable to recognize the distinction; he positively asserts that the distinction does not matter when in a situation of deficient aggregate demand. A lot of the unemployment is in housing and related industries (~2 million jobs lost in construction), but a lot of it is also not in housing and related industries (~8 million jobs lost total). Every industry save health and education has lost jobs since 2008. The expected effect from a large loss of jobs in one sector, absent the resulting loss in AD, would be an increase of employment in other sectors, since the employees freed from the bad sector would lower the price of labor in other sectors. This is simple supply and demand. The fact that the opposite impact is seen every single time strongly suggests that the adjustment is blocked, per Keynes theory.
    - yummyfajitas 16y ago
      That's exactly what I was saying. Keynes assumes he can reduce the world to Calc 101. Austrians/Chicagoans/etc don't. As for why I think it's clear this reduction is not always valid, a simple hypothetical: suppose the stimulus targeted doctor-delivered medicine. There is no unemployment among doctors, so there is no pool of underutilized doctors to employ. It also takes about 8 years to train a doctor, so any effect that stimulus has on inducing construction workers (or others) to become doctors can only occur 8 years later. Also, sticky wages don't immediately lead to the validity of AD or keynesianism. For example, suppose construction workers earned $30/hour before the bust and assume Keynesian wage stickiness. They will not take un-stimulated jobs at Walmart at $12/hour or even stimulated jobs at $24/hour and they will be unqualified for Sous Chef jobs at $30/hour or even a stimulated $35/hour. Unless there is some specific job out there that pays at least $30/hour and construction workers are qualified for it, they will simply remain unemployed. Thus, under some circumstances, Keynes own assumptions make his Calc 101 reduction invalid. Lastly, there is no reason a housing bubble bursting would cause job losses only in construction. It would also harm realtors, mortgage brokers, bankers, etc. It would indirectly harm people who sell things to realtors, mortgage brokers and bankers. Thus, in the recalculation picture, we'd expect the biggest job losses in construction, real estate and mortgages, and correspondingly smaller job losses as you move outward through the economy from this epicenter.