But Keynes Is Saved by Walras's Law

Tyler Cowen writes:

Keynes's General Theory, chapter six: in part ii the bombshell comes, unannounced.  Keynes decides that he will declare savings to be a "mere residual."  Consumption and investment alone will determine income and savings is defined as whatever is left over to make the national income equations balance. At the time this was considered by many to be an enormous sleight of hand. The Austrian and Swedish traditions focused on the question of whether planned savings was going to equal planned investment and what happens if not.  Keynes has just banished such questions to the woodshed and he has done so by a terminological maneuver.

Whether or not you think that the Austrian and Swedish traditions lead anywhere fruitful, Keynes is on shaky ground here.  He is using definitions to favor one causal account of macro over another.  That's not right.  You can still make a plausible argument that Keynes is right on empirical grounds that planned savings is not an important force for understanding business cycles.  But so far no such empirical argument has been clinched...

I think it is much more than a terminological maneuver. Walras's law tells us that if one market is out of supply-demand balance, there must be another related market (or markets) that is also out of balance. If planned saving is in excess of planned investment, then planned consumption spending must be less than planned production of consumer goods. You can then follow the inventory adjustment chain--say that as inventories pile up producers cut back on the making of consumption goods. You then try to follow through on what is happening in the money market and you are led to the conclusion that ex ante savings must be destroyed by a process of deleveraging and deflation and... you wind up in the swamp. But you can be rescued from the swamp by recalling Walras's law, and recognizing that if you just follow the process by which equilibrium is restored in the goods market you will then discover that that process has also restored equilibrium in the flow-of-funds through financial markets.

Keynes's "terminological maneuver" would not have succeeded if it were not for the fact that Keynes's theory worked at a level that Wicksell's or Myrdal's or Ohlin's never could.

Comments

[geoff]

I suspect that if John Maynard wasn't enjoying himself with his Bloomsbury pals somewhere he'd be enjoying the modern, can I call this time period that?, struggles with his clear 1930 thoughts.

[MobiusKlein]

Isn't the distinction in question the difference between planned and un-planned savings?

As an example for the consumer side, candy-canes are selling cheap right now, and any large excess goes to the landfill.
So un-planned savings may well go into landfill worth securities, rather than something actually useful. E.G. all the pets.com style startups that had a domain name and an IPO.

[bakho]

True, it is a definitional part of Keynes model. However, by necessity, models are simplifications of the real world. To be truly helpful, Tyler Cowen needs to explain why defining savings in this manner is problematic and why different models do a better job.

Unless people are hoarding money (literally stuffing mattresses) then the lines between savings, accumulating inventory and investment are blurry.

[Ken Houghton]

Short version of this post: what I don't spend, I save.

Short version of Keynes's "shaky ground": It is easier, in real-time, to measure spending than savings.

Extension: by proxy, a change in inventories represents the difference between expected and actual sales, and therefore (with a sign change) the difference between planned and actual savings.

Weirdest extension that ultimately works: if you consider the inventory change to be "bought" (sold) by the firm, then everything produced is still sold. At which point the inventory change needs to be valued. If you want the math to work (demand=supply), the inventory is then marked at the current sale price.

Extension of the weird extension: Since the firm plans to sell its inventory, it must do so at a market-clearing price. Therefore, the current value of the complete inventory is its Market Value. Conclusion: it must be "marked-to-market" for any reasonable Inventory Management procedure to work.

If this be "shaky ground," UCB has already "tumbled into the sea."

[the idler]

"But Keynes Is Saved by Walras's Law" Tee-hee. Savor that irony, Tyler Cowen!

[MattYoung]

Walrus Law assumes the system in equilibrium.

When we retain money for future consumption we contract the overall economy, that is we cause some producers to drop out.

When we save to fund planned investment, then things work out.

The difference is that the former is an attempt to work more now and less later, which is volatile and economically contracting.

[David Johnston]

When are you economists going to admit that you just make this stuff up and that it is all goobly-gook. Physicists have already admitted it. The game is up.