05
Dec
2012
You Might Be Talkin to a Market Monetarist If…
…you ask this clarifying question on his blog and you genuinely don’t know what answer to expect:
Thanks Nick. One more from me please. (And I’m not asking these to trap you, I’m asking so I fairly recapitulate what your position is.) If the Fed were to suddenly dump its mortgage-backed securities and replace them with gold, would that have any impact on the real estate and mortgage industries, and the world price of gold, relative to the counterfactual? Again, list any caveats you need.
People often say to me, “Bob, for an economist you’re pretty funny.” No I’m not. It’s just that you have to be a trained economist, to recognize how freaking hilarious economists are.
re: ” It’s just that you have to be a trained economist, to recognize how freaking hilarious economists are.”
Definitely agree.
On a somewhat related point I am dumbfounded when people say that Keynes’s writing is “dry”. I think it’s hilarious, or when it’s not outright hilarious it’s at least so clever that it just forces you to smile as you read it. Some of the reason why they say that is because it’s almost a hundred years old now, but I think some of it is that you have to understand the economics involved to get it.
I also think it is hilarious, but more for the same reason why this sort of thing is, as well: http://www.youtube.com/watch?v=WuGsRmEEx6s&feature=g-all
Humor is relative. I don’t think it’s a question of “understanding economics” or not. That just has a nasty ivory tower mentality about it.
Personally, I would understand it if one said they could not smile or chuckle at the writings of an economist who is intellectually responsible for a lot of human suffering, which has lasted almost 80 years now. And I would understand it even if they knew very little economics.
Personally, I have to laugh. It’s a coping mechanism for dealing with the suffering I see around me that is caused by various flavors of statism.
I’ve got a post up on all this that you may be interested in: http://www.factsandotherstubbornthings.blogspot.com/2012/12/on-cantillon-effects.html
I of course come down on Sumner and Rowe when it comes to policy implications of all this, but strictly theoretically speaking I think they oversimplify the Austrian position (which does make a fair amount of sense). Part of that is Sheldon Richman’s fault for focusing a lot on the actual “point of injection” point, when I think Hayek was making a much bigger point than just that.
Current has some important thoughts in the comment section too.
*come down on Sumner and Rowe’s side, I should say.
Then you actually don’t come down on the side of Sumner. He denies it exists….for t-bond monetization.
At any rate, the Cantillon Effect is not an empirically falsifiable proposition. We cannot observe counter-factuals. The Cantillon Effect does NOT say that the first receivers of new money will “receive a rising income trend over time”.
The Cantillon Effect is actually a theoretical construct, that allows us to understand what happens with inflation, as compared to the unobservable counter-factual.
Whatever happens to incomes over time, does not say anything about whether the Cantillon Effect takes place or not.
It doesn’t matter if the initial receiver’s incomes fall over time, or rise over time after an initial inflation injection. If income falls over time, then the Cantillon Effect tells us it fell less than it otherwise would have fell. If it rises over time, then it tells us it rose by more than it otherwise would have risen.
re: “Then you actually don’t come down on the side of Sumner. He denies it exists….for t-bond monetization.”
Ummm… right. That’s exactly why I distinguished between agreeing with them on policy vs. agreeing with them on empirics.
Just FYI, “policy” held constant, i.e. government spending held constant, and changing the locations of governmental money injections, which has the “surprising” result of…no change to “policy”, i.e. no change to government spending….isn’t really a response to whether or not there are relative price and spending changes associated with inflation injections.
OK great, you’re agreeing with them on the assumption of unchanged government spending. But seriously, what’s the importance of that when I thought we were debating Cantillon Effects originating in the t-bond market?
Government spending held constant is not policy held constant IMO. It doesn’t take very long for that to be a quite extreme austerity position. You are really moving the goal posts on policy here.
Sumner is saying that policy held constant is spending held constant.
So what does policy held constant mean to you?
OK, now I don’t know if I understand anything in that comment of yours.
Daniel, you better copyright that. I have a feeling we’re going to see it quoted here a lot.
Wow Ken B, you must have a bone to pick with DK, insulting him like that.
re: “At any rate, the Cantillon Effect is not an empirically falsifiable proposition. We cannot observe counter-factuals.”
I think you mean “not an experimentally falsifiable proposition”.
Or if you don’t mean that you’re wrong and you ought to mean that.
I don’t see the difference between empirically falsifiable and experimentally falsifiable.
In my judgment, falsifiable propositions excludes a priori propositions, and since the only alternative to a priori propositions are a posteriori propositions, it follows that a posteriori propositions are grounded in observation, i.e. empiricism, not a priori reflection/deduction.
Or, another way of saying it: Falsifiable “experiments” are conducted empirically, for the only propositions subject to falsification are empirical ones.
Or maybe you and I are using the word “falsify” differently? I distinguish between falsifiable propositions and refutable propositions, on the grounds that the former are made invalid via observation, and the latter through logic. If that helps?
I am saying that most empirical falsifying work is non-experimental.
Not sure how that’s possible. Maybe you can explain? How can one engage in empirically falsifying a theory, without conducting observations, data collection, and conclusion making (i.e. experimentation)?
Observation, data collection, and conclusion making is not necessarily experimentation.
Don’t worry about it. I’m not sure it’s worth getting into it. I thought you were making a stronger statement about non-experimental empirical work.
What is necessarily experimentation then? Not trying to play gotcha, I just have no idea what you mean by experimentation. I thought it was straightforward.
And DK punts.
Excuse me, “not worth getting into.”
“We cannot observe counter-factuals”
Sometimes we can. That’s the whole idea of controlled experiments.
Counter-factuals are unobservable by definition. It’s why they’re called counter-factuals.
Controlled experiments only reveal factuals, and it’s up to the scientist to infer the causal relation.
Yikes… I agree with MF and disagree with Ken B on this one. This is painful.
Accepting 2+2=4….not painful.
Accepting 2+2=4 in the presence of Hitler accepting 2+2=4….painful.
Ergo, the feeling that comes with thinking true propositions depends on who else agrees or disagrees with them.
Ergo, popularity matters more than knowledge.
Thanks, DK.
Actually what’s painful is acknowledging that Hitler is enough of a rational human to get that 2+2=4.
I’ll let you do the legwork transferring the analogy to our circumstances here.
I thought the analogy was self-explanatory.
OK, here’s an example, tell me why you disagree.
Two large groups of patients, randomly partitioned, all of whom have disease X.
Group A takes a new drug D. All get better, none in the other group do.
This is repeated 17 times, always with the same result.
The control group is a good proxy for the counterfactual, ‘had memebers of group A not taken D they’d still be sick.’ We’re trying to evaluate whether to take D for disease X. Anyone not convinced it’s a good drug to take?
Glad you agree that counter-factuals are not observable. That you are compelled to use observable “proxies” that you infer as what otherwise would have occurred.
While this method is statistically successful, it is not perfect, after all, the 18th trial may kill everyone, but more important than that, the inference you are making is based on observing factuals and thinking, not observing, counter-factuals.
Oh, if you are making that point yes. That’s trivially true. I thought you were arguing we could not sensibly have or update priors about such things.
It’s like ifsomeone said, had the south won the civil war they would have boiled all the blacks in pots and feasted on their flesh. That’s an unobservable counter factual but I expect we all evaluate the probability at close to 0.
Actually I was just disputing this point you made:
I said: “We cannot observe counter-factuals”
You said: “Sometimes we can.”
If you can see how that is wrong, if you can see that no, we can NEVER observe counter-factuals, not even “sometimes”, then I think we can move on.
Daniel, I didn’t mean to overstress the point of injection. I just wanted to say that the money has to start somewhere. The article was about much more than this.
Definitely Sheldon. And “due to” would probably have been better word choice on my part than “fault”.
Even if you had been talking all about injection points, it can be difficult to anticipate in what direction a particular discussion would go in. I just did my post because I think we could generalize your (in my view theoretically valid but practically less important) point beyond the injections themselves.
Why not inject all new money via direct payment to producers below 150K and let prosperity trickle up rather than poverty trickle down?
Could this really be Sumner subconsciously defending NGDP theory from the rabble, who would “fiscally” revolt if they knew the Cantillon Effect applied to “market monetarism” t-bond monetization no less than gold or car purchases?
Is that why he wants to say it’s “fiscal”, i.e. “public”, i.e. “political”, when the Fed buys something other than t-bonds from the market, but it is “monetary” when it buys t-bonds, so whew(!) any wealth redistribution associated with market monetarism is covered up? Again, I am asking if this is his subconscious principles coming to the fore.
” It’s just that you have to be a trained economist, to recognize how freaking hilarious economists are.”
That’s proabaly true of most things. I’m told accountants and actuaries are really hilarious.
Bob: I don’t get the joke!?
Quick answer: holding M and expected future M constant, exactly the same effect as if you or I sold our MBS and bought gold. (OK, I expect the Fed is a bit bigger than you or me). Price of gold up a bit, and price of MBS down a bit.
Watch out Bob. It’s that Canadian dry humour.
OK thanks Nick. I just ask because I want to be crystal clear when I handle Scott’s point that the Fed buying an asset from someone doesn’t help him, because it occurs at the “market price.” If the very act of buying it raises the market price, then that is relevant to the discussion.
OK, I don’t get Nick’s answer maybe. To some extent prices must surely depend on a reading of the Fed’s intent. So suddenly the Fed has a complete change of heart on an asset it has been willing to hold a lot of, in this case MBS. We want none of them, none you hear me, none!
I suspect I am *gulp* agreeing with Bob on this, so Bob’s comment here is (hilarious!!) understatement. Is that right?
Everything except the gulping part.
the Fed buying an asset from someone doesn’t help him, because it occurs at the “market price.
Isn’t that a little like saying that buying a pizza at my local pizzeria doesn’t help Luigi because I buy it at the market price.
But Luigi is just “swapping” a $10 pizza for $10 in money. He isn’t any wealthier. He doesn’t have an incentive to spend more.
/s
No Luigi would rather have the $10 than the Pizza; that why he bothered to do it. So Luigi sees himself as better off; at least that’s his revealed preference.
But Luigi is retarded if that’s his preference. It’s like saying you prefer one $10 bill over another. There is really really really no difference.
Oh, an maybe we should add:
It’s not selling pizzas that makes Luigi wealthier. If any, it’s producing pizzas. As of the moment the pizza is done, his balance sheet grew by x. Trading his assets for a more liquid class of assets is a wash. Luigi couldn’t care less.
Gosh, I can’t believe people still get this wrong…
Except he’s trading $10 for a pizza, not another $10. Unless you were being sarcastic.
Ok Christopher something about your post signals you weren’t being sarcastic.
You claim Luigi becomes wealthier when he produces the pizza, not when he sells the pizza for $10. In other words, you are arguing that “it’s a wash” to sell the pizza. but then why is he selling the pizza in the first place? Obviously it’s because he values the money more than he values the pizza.
By definition that means he is wealthier with the $10 than he is with the pizza. Wealth is determined by the valuations of individuals involved, not your arbitrary opinions about the matter.
Ok Christopher something about your post signals you weren’t being sarcastic.
You claim Luigi becomes wealthier when he produces the pizza, not when he sells the pizza for $10. In other words, you are arguing that “it’s a wash” to sell the pizza. but then why is he selling the pizza in the first place? Obviously it’s because he values the money more than he values the pizza.
By definition that means he is wealthier with the $10 than he is with the pizza. Wealth is determined by the valuations of individuals involved, not your arbitrary opinions about the matter.
I was being sarcastic.
Should have went with my initial suspicion.
I was being sarcastic
It is.
What if M suddenly started to grow at a constant rate, and the entire additional M each period was used to buy gold.
Is it correct to say “The price trend of gold over time will REMAIN elevated as compared to the counter-factual price trend associated with constant M.”?
I am reminded of a recent comment by Bill Gross:
“We continue to anticipate what the Fed is buying,” Gross said. “They’ve told us they will buy $40 billion to $70 billion of agency mortgages every month until the cows come home. It pays to own these mortgages even though they’re overvalued.”
http://www.cnbc.com/id/49180989/We_re_Buying_What_the_Fed_and_ECB_Are_Buying_Gross
Yancey,
I was searching Webster’s for the definition of mal-investment and it brought me your Bill Gross quote, strange.
“Bob, for an economist you’re pretty funny.” No I’m not.
Maybe just a little funny. Name this Zombie:)
http://www.youtube.com/watch?v=TrcM5exDxcc&feature=player_detailpage
Bob, is there some significance to “the Fed” in the question? Presumably you don’t believe the Fed’s transactions have a different effect on markets than, say, Warren Buffett’s – except when they’re done specifically for the purpose of setting the Fed Funds rate.
Max yes there is some significance. Scott and Nick both think the Fed might make bonds cheaper by buying them, because in so doing it creates more money and raises long-run price inflation expectations. That can’t happen with buyers who are unable to create money out of thin air by buying something.
Long run price inflation expectations lowering the current price of bonds might make sense if the inflation arises out of something other than bond purchases.
But if the inflation arises out of buying bonds, then the traditional bond pricing model that includes an inflation premium, does not apply. In these cases, it must be understood that any temporal decline in bond prices are smaller than they otherwise would have been, and so there is still a premium added to the bonds, the counter-factual of which we cannot observe.
The way to get this is to compare the following two scenarios of inflation:
1. The Fed inflates and spends $X money on consumer goods, thus triggering an inflation premium on t-bonds. T-bond prices are lower than they otherwise would have been.
2. The Fed inflates and spends $X money on t-bonds, thus triggering an inflation premium on t-bonds. T-bond prices are lower than they otherwise would have been.
Question: Are the resulting prices of t-bonds identical in these two scenarios? If we are to believe Sumner and Rowe, then the answer is yes. If however we are to think that the decline in t-bond prices in the second scenario is less than in the first scenario, then we must reject the position of Sumner and Rowe and re-adopt the very same economic science that applies to Fed purchases of cars or gold.
Ok, but Scott would say that the purchase is all about communication. And the Fed could just as well communicate using English. In fact, that would be better.
Bob, are you writing a detailed reply to all of this? Someone has to. This seems to me to be pretty basic, yet its going right over their heads.
Yes I am Chris, but it’s for an outlet that might not run it for a month or more, so pace yourself…