Saturday, April 4, 2015

Macroeconomics = Fallacy of Composition

I am teaching an introductory macroeconomics course for graduates this semester at the University of Queensland. Coming up is soon is the AD-AS model and I must say, I’m having a hard time generating useful tutorial questions for my students.

The illogical contrivances required to interpret any real economic events in terms of the model are almost laughable. I've asked a number of experienced macro teachers in the school to explain the model to me in an internally consistent way. It turns out that most economic professors don’t have a clear grasp of what aspects of the macroeconomy the model is meant to be capturing. None could explain what the concept of the price level in the aggregate even is, nor what mechanism was meant to be at play in generating the relationship between price level and output. If anything I learnt that I should use the curves to support some kind of Keynesian story-telling exercise.

What kind of serious intellectual could be happy with this situation? Is this the way we pass on our discipline’s knowledge to the next generation?

So far my tutorials are looking quite different from your typical introduction to the AD-AS model. More like an introduction to defending oneself against the claptrap sprouted by many economists.

Perhaps we may soon leave this nonsense out of economic teaching altogether. Here's my compromise for the time being.

Question 1

What is the fallacy of composition?

Create a list of phrases, sentences or paragraphs from the description of the AD-AS model in Chapter 10 of the textbook that commit the fallacy of composition.

Discussion

A common definition for the Fallacy of Composition is
the fallacy of inferring that a property of parts or members of a whole is also a property of the whole
In terms of the study of macroeconomics this is of particular interest, since we are trying to take lessons from the behaviours of individuals and determine the likely aggregate response in the economy as a whole.

Here are some examples from Chapter 10, though many others may be included.

1. The aggregate demand curve shows the relationship between the price level and the quantity of real GDP demanded by households (p268).

The AD curve is downward sloping because a fall in the price level increases the quantity of real GDP demanded (p268)

The fallacy at play here is that there is an aggregate price level. As we saw in Chapter 8 on Inflation, price levels are not absolute but relative to some base year. In the economy as a whole there is no external price from which to determine a price level. Hence the idea of an economy-wide price level is in a given time period is a fallacy.   Often the AD-AS model is interpreted as showing Real GDP related to the rate of inflation in a period (not the level, but the rate of change of price levels relative to a base year). This overcomes the fallacy of composition because it compares the price level with last year’s price level, but contradicts earlier discussions about anticipated vs unanticipated inflation. If inflation is anticipated at any reasonable level then there should be no economic effect, and hence no relationship (no curve, or a vertical line at best).

2. As income rises consumption will rise, and as income falls, consumption will fall (p269)

In aggregate the total consumption (or demand) in the economy is equal to the total incomes in the economy be definition, since someone’s consumption is somebody else’s income. This sentence, allowing from proper aggregation that avoids the fallacy of composition, merely says that aggregate incomes and consumption just rise and fall together since they are equal at any point in time by definition.

3. When prices rise, households and firms need more funds to finance buying and selling. Therefore when the price level rises, household and firms try to increase the amount of funds they hold by withdrawing from banks, borrowing from banks or selling financial assets, such as bonds. These actions tend to drive up the interest rate charged on bank loans and the interest rate on bonds. (p269)

First, if all prices rise then households and firms also have more income, since wages and profits are also prices in the economy.

Second, as we discussed in Chapter 4, Measuring GDP, transfers of financial assets are not included in gross value added because they are mere transfers between on individual and another. Hence, a household selling financial assets requires another household to buy those financial assets leading to no net change in the aggregate ability of households to spend on current consumption.

Third, if these actions have no aggregate affect by their own definitions it seems implausible that there will be aggregate affects on interest-rates (another price) set in the economy.

The end of the paragraph goes on to admit that it just committed the fallacy of composition when it says
The impact of the price level on investment is known as the interest-rate effect, and is a second reason why the AD curve is downward sloping. However there is caveat to this explanation. Lenders to banks and other financial institutions - people with savings are lenders - will have their wealth increased as the interest rate rises, and will increase their consumption spending.
Essentially this contradicts all that went before it. In addition, the wealth effect on savers is incorrect in general (higher interest rates decrease the value of financial assets, not increase it).

4. Table 10.1, government purchases and tax effect on AD curve (p272).

The second and third rows in Table 10.1 show; 1) the effect of an increase in government purchases on, 2) the effect of an increase in government taxation. If the economy is at equilibrium prior to and after either of these changes then by definition one effect must counteract the other - the shift of the AD curve from an increase in government purchases must be accompanied by an equal and opposite shift from an increase in taxation.

5. Because changes in the price level do not affect the number of workers, the capital stock or technology, in the long run changes in the price level do not affect the level of real GDP (p275).

This contradicts all that went before it, and relies on an undefined notion of the short and long run. How can there simultaneously be both an effect and no effect?

6. The main reason firms are willing to supply more goods and services as the price level rises is that, as prices of final goods and services rise, prices of inputs - such as wages of workers to the price of natural resources - rise more slowly. Profits rise when the prices of the goods and services firms sell rise more rapidly than the prices they pay for inputs (p276).

Here we see the fallacy of composition result in many confusing comments. First, the inputs of one firm are the outputs of another firm, hence in aggregate input and output prices cannot differ systematically.

1. "Create a list of phrases, sentences or paragraphs from the description of the AD-AS model in Chapter 10 of the textbook that commit the fallacy of composition."

What textbook is that?

1. http://www.pearson.com.au/products/H-J-Hubbard-Glenn-Et-Al/Macroeconomics/9781486010233?R=9781486010233

2. "What kind of serious intellectual could be happy with this situation?"

The kind that is motivated by political ideology rather than a genuine interest in the open-minded pursuit of demonstrable, reliable truth. And that includes all the Keynesians and the monetarists and pretty much all the other macro-guys.

'Death by torture to all macro-witchdoctors.'

3. Excellent

4. I really don’t like to blow my own horn, but for these issues you should see my textbook, Macroeconomics: A Fresh Start. The fallacy of composition problem is front and center throughout, and AS-AD and IS-LM are both treated honestly. I’m curious to know what you would think.

1. If you don't blow that horn, no on will listen. Let me get a copy. Thanks

5. Hi Cameron,

You inspired me to look more closely at the AD-AS model. I'm actually unfazed the fallacy of composition -- I think it would be better named the "warning of composition" because in many cases 1 + 1 + 1 ... + 1 really is just N. Sometimes it's not (the SMD theorem comes to mind, and nominal rigidity seems to scream out as an effect that doesn't exist at the micro level), so one should be careful ...

1. To be clear, the fallacy of composition arises in the first instance in the case of the price level. For there to be a price level, we are comparing the prices of goods in terms of the unit of account. Now exactly how can that level have some intrinsic relationships with output? After all, the units of money accounts are completely arbitrary. We must have in mind not a price level (which is a complete arbitrary construction in any given year), but a change in the price level over the period which we measure output. This is exactly the modification that Romer makes if you follow soleus's suggestion in the comments.

The second big example of the fallacy is in terms of the interpretation of rising incomes. In the aggregate the sum of all incomes equals the sum of all spending. Saying 'as incomes rise consumption will rise' tells us nothing, because by definition the aggregate income and consumption must be equal and hence rise and fall in tandem.

I've read your post but am honestly a bit confused about how your reasoning relates to the fallacy of composition. The fallacy arises when trying to construct an AD or AS curve, not in the subsequent manipulations of them.

2. Hi Cameron,

I guess I wrote up the AD-AS model post a bit too quickly and really only addressed the fallacy of composition with a single sentence:

"[the model] essentially lives entirely on the macro scale, so there isn't any fallacy of composition."

That is to say the model makes no specific assumptions about the details of household or firm behavior. You could interpret that in a couple of ways, though.

1. Micro behavior is approximately random
2. Micro behavior is so complex it appears random
3. Any detailed micro behavior is subject to the fallacy of composition -- after being filtered through the aggregation process, all that's left looks random

'Random' here is being used in the thermodynamics sense -- I'm not saying households are irrational, increasing or decreasing their holdings of liquid assets at random month to month. Each atom in a gas obeys very strict physical laws like momentum and energy conservation. Likewise each economic agent could have some kind of deterministic micro behavior. Just like we don't know the history of collisions for an atom in a gas, we don't know the history of financial transactions of one household. Any given household will be increasing or decreasing its holdings of liquid assets at any given time because of that history. Most of the time these are in "detailed balance": households increasing equals households decreasing.

If e.g. inflation falls, there could be a tendency on average for those holdings to increase (based on millions of household decisions consistent with millions of financial histories) due to a small imbalance, aggregated up to the macro level. However, any individual household wouldn't point to the falling inflation. And it's not even true that inflation is influencing household behavior -- falling inflation is simply the most likely macro state where liquid asset holdings are increasing (in the AD-AS model). Since there are millions of households, the law of large numbers kicks in.

In that way, the AD-AS model I present at the link is a macro-only model. Higher inflation doesn't cause individual households to increase their demand for iPads. The higher inflation macro state is consistent (in the AD-AS model) with a macro state in which more iPads are being consumed.

Sorry for the long comment.

6. I see the AS/AD model as mainly useful for showing how the fiscal policy response to lowering interest rates (does it increase investment in projects whose NPV has become positive at the new interest rates or doe it opt for "austerity") works out.

1. Can you briefly comment on how the model can be used in this way? Because I can't see it.

7. Get hold of David Romer's 2013 lecture notes on macro, easy to find online. He fixes the worst problems with AS/AD.

1. Maybe he fixes the worst (by using inflation rather than price level), but to be honest, there is nothing in that document that tells me anything meaningful about reality. There is merely a lot of reverse engineering an explanations for observed events into the model, and not discussion of mechanisms (eg. what on earth the inflation adjustment line is meant to be).

To be honest, this only reinforces my concern about the training of economists. How can most textbooks say the AD-AS model relates the price level to output, and another the inflation rate? Imagine in physics if some textbooks said energy was half mass times position squared, and other said half mass times velocity? No one would take it seriously. And nor should they. Yet this is exactly the situation in macro economics.

8. Not SO easy. You need to google 'Romer short run fluctuations'

9. Get a better textbook. The author of yours is highly flawed.

http://krugman.blogs.nytimes.com/?s=hubbard

10. Some guy once wrote:
"The well-known, but unavoidable, element of vagueness which admittedly attends the concept of the general price-level makes this term very unsatisfactory for the purposes of a causal analysis....To say that net output to-day is greater, but the price-level lower, than ten years ago or one year ago, is a proposition of a similar character to the statement that Queen Victoria was a better queen but not a happier woman than Queen Elizabeth — a proposition not without meaning and not without interest, but unsuitable as material for the differential calculus. Our precision will be a mock precision if we try to use such partly vague and non-quantitative concepts as the basis of a quantitative analysis."
That was Keynes:
https://www.marxists.org/reference/subject/economics/keynes/general-theory/ch04.htm

11. Had to check whether this was April 1st post, but apparently not... since all these arguments against "fallacies" of AS/AD model are in fact themselves incorrect. So here's the debunking of the debunking:

1. "The fallacy at play here is that there is an aggregate price level. As we saw in Chapter 8 on Inflation, price levels are not absolute but relative to some base year"

So? Obviously the whole model is worked out assuming some particular base year. Choosing a different base year would merely rescale the P axis, and bear no relevance to qualitative predictions of the model.

Alternatively, and for greater internal consistency, we can think of the model as a stylized example of an artificial economy with four goods (output, money/currency, bonds, labor), money chosen as numeraire and price level is simply the (well-defined) price of the single output good.

2. "In aggregate the total consumption (or demand) in the economy is equal to the total incomes in the economy be definition, since someone’s consumption is somebody else’s income."

The book is likely speaking about private consumption expenditures, not the whole aggregate demand. Aggregate consumption function C = a+b*Y is pretty standard, is it not?

Moreover, even if the model includes one relationship between two variables, it can include also another relationship, with both being true simultaneously. Like you know, the supply and demand model has both positive and negative relationship between price and quantity.

3. "First, if all prices rise then households and firms also have more income, since wages and profits are also prices in the economy."

Nominal incomes rise. Real incomes don't have to.

"Second, as we discussed in Chapter 4, Measuring GDP, transfers of financial assets are not included in gross value added because they are mere transfers between on individual and another."

Transfers have nothing to do with this. Here all the households find themselves with too little money (real money balances have decreased), thus they all prefer to adjust their portfolios towards holding currency. If the CB doesn't adjust the money supply, equilibrium must be restored by decrease in bond prices, i.e. higher interest rate.

4. "the shift of the AD curve from an increase in government purchases must be accompanied by an equal and opposite shift from an increase in taxation."

No. First, the budget doesn't have to be balanced. Even if it is, the two shocks don't have to have exactly same countervailing effects (exercise: derive balanced budget multiplier in the Keynesian cross model).

5. "This contradicts all that went before it, and relies on an undefined notion of the short and long run. How can there simultaneously be both an effect and no effect?"

On impact, there is some effect. In subsequent periods, the effect will start to dissipate, and eventually the economy will converge back to the original equilibrium. It's not so hard to get.

6. "Here we see the fallacy of composition result in many confusing comments. First, the inputs of one firm are the outputs of another firm, hence in aggregate input and output prices cannot differ systematically."

The example talks about specifically about rigid wages. Last time I checked, labor is not produced as output by any firm.

12. "turns out that most economic professors don’t have grasp of what aspects of the macroeconomy the model in meant to be capturing"

Why am I not in the least surprised...

13. I don't see it as a fallacy that macroeconomics as a whole must take into account feedback that microeconomics can ignore. That just says that the aggregate supply and demand curves don't explain the whole picture, which you can note after discussing what the graph is supposed to represent. And I don't see what's wrong with telling students that in the current real world, aggregate price levels are measured in terms of fiat currency units.

I don't know which way around you put the axes, but I will do my best to suggest a couple of useful examples of the use of the aggregate supply and demand curves without using terms like "down" and "up".

Example number one: traditional Keynesian demand side stimulus, which is an increase in central government spending, fully monetized such that it results in an increase in nominal debt and no increase in nominal taxation. This moves the demand curve in the direction of increasing quantity. The net effect is that the equilibrium quantity is higher than it was initially, by less than the amount of the spending, and the equilibrium price is also higher than it was initially.

Point out that amounts of the changes in quantity and price depend on the local slopes of the demand and supply curves. Relate this to 1970s inflation.

The discussion to this point should ignore feedback effects. At this point, though, you can point out that the feedback effects may exist. You can talk about how there may be a multiplier effect if the sellers to the government use the money to increase their own demand, resulting in a "multiplier" effect under some conditions. You can talk about how inflation expectations may change the effective shape of the supply curve and maybe the demand curve. You should definitely talk about how feedback effects mean that the aggregate supply and demand model, while useful, is not a complete picture.

Example number two: idealized supply side stimulus. For example, the government could reduce income tax rates, effectively reducing the price of goods, again monetizing any associated debt. This moves the supply curve in the direction of decreased prices. This also results in an increase in equilibrium quantity, but this time with a decrease, rather than an increase, in nominal prices. Discuss how supply side policies were used effectively in the US in the 1980s in a high inflation environment where deflationary pressure was beneficial, and how it caused the government to rack up a lot of debt, but without driving interest payments higher since falling inflation resulted in neutral to falling nominal interest rates.

Again, after presenting the example, you can discuss feedback effects.

Example number three: monetary stimulus. Note how it's difficult to see how this relates to the aggregate supply and demand graph. Does it just change the units on the price axis? Note how this is why macroeconomics needs a lot more than the aggregate supply and demand curve, and how a lot of macroeconomics is about nonequilibrium manipulation, for which an equilibrium model like the supply and demand model may be less than useful.