• Tweet

• Tweet

Jet­lag has me up and at the key­board at 5.54am here in Lon­don, 43 min­utes before sun­rise, which today is at 6.37am.

Only it’s not “sun­rise”, is it? As we all know, it’s real­ly “Earth Axi­al Rotate” at the point in its 24 hour axi­al rota­tion when the Sun—around which the Earth rotates once each year—becomes vis­i­ble from Lon­don.

Click here for this post in PDF

We still call it “sun­rise” because it’s a lot less awkward—and a lot more romantic—than say­ing “Earth Axi­al Rotate Earth-Sun Radi­al Align­ment”, which is what it real­ly is. We all know that it’s not real­ly the Sun “ris­ing” at all: that implies that the Earth is fixed while the Sun rotates around it, where­as ever since Coper­ni­cus we have known that, though it looks that way to a naïve observ­er on Earth, that’s not what real­ly hap­pens.

How­ev­er, not mere­ly before Coper­ni­cus, but for a very long time after him, many peo­ple con­tin­ued to believe that that was how it real­ly is: that the Sun does rotate around the Earth, that the Earth is not mere­ly fixed, but fixed at the Cen­tre of the Uni­verse, and not mere­ly the Sun but all Celes­tial bod­ies rotate around it in per­fect spheres.

What broke us from that belief was the empir­i­cal fail­ure of the the­o­ry which encap­su­lat­ed it and still made sense—as much as it could—of the anom­alies between the pre­dic­tions of that the­o­ry, and actu­al real­i­ty. Claudius Ptole­my’s trea­tise the “Math­e­matike Syn­taxis” (or Math­e­mat­i­cal Com­po­si­tion), which became known as the Almagest (mean­ing “The Great Trea­tise”), was pub­lished in about 150 BC, and it pro­vid­ed a plau­si­ble mod­el for earth-cen­tric beliefs about the nature of the Uni­verse that dat­ed back mil­len­nia. It held sway not mere­ly until Coper­ni­cus wrote his De rev­o­lu­tion­ibus orbium coelestium in 1543, but for many decades after, as not only the Church but also incensed Ptole­ma­ic astronomers fought to sup­press the new, more accu­rate, but to them hereti­cal and false mod­el of the Uni­verse.

Why the brief dis­course on Astron­o­my? Because read­ing what Paul Krug­man is say­ing about bank­ing feels like read­ing a Ptole­ma­ic Astronomer describ­ing sun­rise today as if that’s actu­al­ly what’s hap­pen­ing. He is dis­mis­sive of the view that banks can “cre­ate cred­it out of thin air”—so dis­mis­sive in fact, that any­one unac­quaint­ed with the empir­i­cal evi­dence might be fooled into believ­ing that his case is so strong­ly sup­port­ed by the facts that it’s not even worth the both­er of cit­ing the empir­i­cal data that backs it up.

That is so NOT the case: the empir­i­cal evi­dence over­whelm­ing­ly sup­ports the case Krug­man is try­ing to dis­miss out of hand, that banks can and do “cre­ate cred­it out of thin air”, with the sup­posed reg­u­la­to­ry con­trols over their capac­i­ty to do so being large­ly inef­fec­tive.

Fig­ure 1: Krug­man’s 3rd post on bank­ing and mon­ey cre­ation

In fact the evi­dence is so strong­ly in favour of the case that Krug­man blithe­ly dis­miss­es that it’s dif­fi­cult to decide where to begin in refut­ing his Ptole­ma­ic fan­tasies to the con­trary. I’ll lead with his “gotcha!” argu­ment in this post, but before that I’ll return to the Ptole­ma­ic Astron­o­my-Neo­clas­si­cal Eco­nom­ics analogy—because it’s quite a strong one that deserves fur­ther elu­ci­da­tion.

Ptolemy and Walras—Brothers in Arcs

The Geo­cen­tric mod­els of the uni­verse, of which Ptole­my’s sys­tem was a vari­a­tion, had 3 guid­ing prin­ci­ples, which Cardall describes as fol­lows:

1. All motion in the heav­ens is uni­form cir­cu­lar motion.
2. The objects in the heav­ens are made from per­fect mate­r­i­al, and can­not change their intrin­sic prop­er­ties (e.g., their bright­ness).
3. The Earth is at the cen­ter of the Uni­verse. (Cardall, 2000)

The key prob­lem with this base the­o­ry is that it man­i­fest­ly did­n’t fit the facts, because of the behav­iour of celes­tial bod­ies that we now call Planets—which is the ancient Greek word for “wan­der­ers”. Far from obey­ing uni­form cir­cu­lar motion, these Wan­der­ers lit­er­al­ly did wan­der all over the sky. We’re gen­er­al­ly not aware of this today because it’s no big deal from our bet­ter-informed Helio­cen­tric mod­el of the solar sys­tem, but for the ancients it was a big deal. A sim­u­la­tion by David Colarus­so indi­cates how much the appar­ent behav­iour of the Wan­der­ers vio­lat­ed the three core tenets of the Geo­cen­tric mod­el.

Ptole­my’s con­tri­bu­tion was to pro­vide “tweaks” to this core vision, which main­tained its over­all integri­ty while fit­ting it much more close­ly to the data. He stuck with most of propo­si­tion (1) and all of (2), but mod­i­fied (3) to “The Earth is near the cen­ter of the Uni­verse”. With the Earth slight­ly off-cen­ter, the gen­er­al­ly ellip­ti­cal motion of The Wan­der­ers could now be explained by what was called The Eccen­tric. But their habit of “ret­ro­grade” motion—the fact that they would occa­sion­al­ly reverse direc­tion in the night sky—was still an anom­aly.

To solve that, Ptole­my added cir­cu­lar motion on cir­cu­lar motion. All celes­tial bod­ies still fol­lowed a great circle—called the Deferent—but the plan­ets also did their own rota­tions on the Def­er­ent on mini-cir­cles called Epicy­cles.

But even that was­n’t enough, because the plan­ets also appeared to speed up on part of their motion through the heav­ens, and slow down on oth­ers (today we know this is just because some­times they are clos­er to the earth on their ellip­ti­cal orbits around the earth, and there­fore appear to move more rapid­ly). So Ptole­my added “Equant” motion: the big “Def­er­ent” cir­cle each plan­et moved on was divid­ed into seg­ments by lines through a point which was not its cen­ter, and the plan­et moved through each dif­fer­ent­ly sized slice in the same time—thus speed­ing up in the big slices and slow­ing down in the small ones.

By these tweaks, a par­a­digm which was utter­ly unlike the real world was actu­al­ly able to mim­ic it to a tol­er­a­ble lev­el of accu­ra­cy. But the sys­tem was extreme­ly com­pli­cat­ed, and it took an enor­mous amount of brain pow­er to be a Ptole­ma­ic astronomer. Look­ing back on this once dom­i­nant the­o­ry, Cardall telling­ly observes how the very com­plex­i­ty of this absolute­ly false men­tal con­struct helped pre­serve it despite mount­ing evi­dence that it did not describe real­i­ty:

That ancient astronomers could con­vince them­selves that this elab­o­rate scheme still cor­re­spond­ed to “uni­form cir­cu­lar motion” is tes­ta­ment to the pow­er of three ideas that we now know to be com­plete­ly wrong, but that were so ingrained in the astronomers of an ear­li­er age that they were essen­tial­ly nev­er ques­tioned. (Cardall, 2000)

Why am I remind­ed of Neo­clas­si­cal Eco­nom­ics? Let me count the ways…

First­ly, there are sim­i­lar under­ly­ing prin­ci­ples to the DSGE mod­els that now dom­i­nate Neo­clas­si­cal macro­eco­nom­ics, and as with Ptole­ma­ic Astron­o­my, these under­ly­ing prin­ci­ples clear­ly fail to describe the real world. They are:

1. All mar­kets are barter sys­tems which are in equi­lib­ri­um at all times in the absence of exoge­nous shocks—even dur­ing recessions—and after a shock they will rapid­ly return to equi­lib­ri­um via instan­ta­neous adjust­ments to rel­a­tive prices;
2. The pref­er­ences of con­sumers and the tech­nol­o­gy employed by firms are the “deep para­me­ters” of the econ­o­my, which are unal­tered by any poli­cies set by eco­nom­ic pol­i­cy mak­ers; and
3. Per­fect com­pe­ti­tion is uni­ver­sal, ensur­ing that the equi­lib­ri­um described in (1) is social­ly opti­mal.

If that were actu­al­ly the real world, then not only would there not be a cri­sis now, there would nev­er have been a Great Depres­sion either—and reces­sions would sim­ply be minor sta­tis­ti­cal­ly unpre­dictable but inevitable events when the major­i­ty of shocks hit­ting the econ­o­my were neg­a­tive, and they would rapid­ly be resolved by adjust­ments to rel­a­tive prices (wages includ­ed, of course).

So econ­o­mists like Krugman—who describe them­selves as “New Keynesians”—have tweaked the base case to derive mod­els that “ape” real-world data, with “sticky” prices rather than per­fect­ly flex­i­ble ones, “fric­tions” that slow down quan­ti­ty adjust­ments, and imper­fect com­pe­ti­tion to gen­er­ate less-than-opti­mal social out­comes.

This is Ptole­ma­ic Eco­nom­ics: take a mod­el that is utter­ly unlike the real world, and which in its pure form can’t pos­si­bly fit real world data, and then add “imper­fec­tions” so that it can appear to do so.

Fig­ure 2: Krug­man’s 3rd post on bank­ing and mon­ey cre­ation

Walk like a Ptolemain

Krug­man’s rejec­tion of the propo­si­tion that banks can cre­ate money—in the sense that “their abil­i­ty to cre­ate mon­ey is not con­strained by the mon­e­tary base” as he puts it in an update—is also a vin­tage Ptole­main manoeu­vre. A sci­en­tif­ic response to this propo­si­tion would be to dis­prove it via empir­i­cal evi­dence. Krug­man instead appeals to his own author­i­ty, relies on deduc­tive logic—which I’ll return to shortly—and derides those who believe that banks and cred­it growth mat­ter in macro­eco­nom­ics as “Bank­ing Mys­tics”.

What Krug­man dis­plays here is not greater insight but blind igno­rance. A recent addi­tion to the over­whelm­ing evi­dence that cred­it growth is a cru­cial fac­tor in macro­eco­nom­ics is an empir­i­cal paper by those well-known bas­tions of Bank­ing Mys­ti­cism, the Nation­al Bureau of Eco­nom­ic Research and the Fed­er­al Reserve Bank of San Fran­cis­co. The paper analy­ses 200 reces­sions in 14 coun­tries over 140 years, and sum­maris­es its results as fol­lows:

This paper stud­ies the role of lever­age in the busi­ness cycle. Based on a study of near­ly 200 reces­sion episodes in 14 advanced coun­tries between 1870 and 2008, we doc­u­ment a new styl­ized fact of the mod­ern busi­ness cycle: more cred­it-inten­sive booms tend to be fol­lowed by deep­er reces­sions and slow­er recov­er­ies. We find a close rela­tion­ship between the rate of cred­it growth rel­a­tive to GDP in the expan­sion phase and the sever­i­ty of the sub­se­quent reces­sion. We use local pro­jec­tion meth­ods to study how lever­age impacts the behav­ior of key macro­eco­nom­ic vari­ables such as invest­ment, end­ing, inter­est rates, and infla­tion. The effects of lever­age are par­tic­u­lar­ly pro­nounced in reces­sions that coin­cide with finan­cial crises, but are also dis­tinct­ly present in nor­mal cycles. The styl­ized facts we uncov­er lend sup­port to the idea that finan­cial fac­tors play an impor­tant role in the mod­ern busi­ness cycle. (Oscar Jor­da et al., 2011a, Oscar Jor­da et al., 2011b)

That emphat­i­cal­ly decides the key empir­i­cal dispute—whether the lev­el and rate of growth of aggre­gate pri­vate debt has macro­eco­nom­ic effects—in favour of the case I put.

Unreserved Lending

There is also a wealth of stud­ies to sup­port the con­tention that reserves don’t con­strain lending—that if any­thing, the causal link runs from lend­ing to reserves, and not the oth­er way around. I referred to some of these in my last blog post, so I won’t repeat that issue here. Instead I’ll take up Paul’s “gotcha” argu­ment to the con­trary:

Yes, a loan nor­mal­ly gets deposit­ed in anoth­er bank — but the recip­i­ent of the loan can and some­times does quick­ly with­draw the funds, not as a check, but in cur­ren­cy. And cur­ren­cy is in lim­it­ed sup­ply — with the lim­it set by Fed deci­sions. So there is in fact no auto­mat­ic process by which an increase in bank loans pro­duces a suf­fi­cient rise in deposits to back those loans, and a key lim­it­ing fac­tor in the size of bank bal­ance sheets is the amount of mon­e­tary base the Fed cre­ates — even if banks hold no reserves.

Sigh. The lev­el of cur­ren­cy retrains lend­ing? So banks stop lend­ing as they approach the lim­its to cur­ren­cy set by the Fed’s print­ing of notes?

I can’t improve on the com­ments of Neil Wil­son on Krug­man’s argu­ment here:

Krug­man needs to start attend­ing the real world. The lat­est argu­ment is utter tosh. For there to be a con­straint in the real world, you have to have the actu­al pow­er to stop anoth­er enti­ty from doing some­thing.

What Krug­man is sug­gest­ing is that the Fed has the pow­er to lim­it the amount of cur­ren­cy in issue. In oth­er words he’s sug­gest that to con­trol the econ­o­my the ATMs will be left to run dry and you will be told ‘no’ when you go and try and draw cash at the bank counter.

Sweep­stake on how many attosec­onds it would take to cause gen­er­al pan­de­mo­ni­um if that every hap­pened. Here in the UK there has been a sug­ges­tion that the fuel pumps might be short of fuel if the tanker dri­vers did decide to go on strike. It has caused com­plete chaos even though noth­ing is dif­fer­ent this week­end than last. Krug­man is beyond grasp­ing at straws now.

And even if the Fed could do that—even if it did attempt to con­trol bank lend­ing by manip­u­lat­ing reserves (some­thing it gave up on doing about 30 years ago)—there are two fac­tors need­ed to make manip­u­lat­ing reserves a con­trol mech­a­nism over bank lend­ing:

1. Reserves them­selves; and
2. A man­dat­ed ratio between deposits at banks and reserves

Paul does­n’t seem to have caught up with the fact that this man­dat­ed ratio no longer exists, for all prac­ti­cal pur­pos­es, in the USA and much of the rest of the OECD. Six coun­tries have no reserve require­ments what­so­ev­er; the USA still has one, but for house­hold deposits only. Fig­ure 3 shows the actu­al rules for reserves in the USA—taken from an OECD paper in 2007 (Yueh-Yun June C. O’Brien, 2007). The reserve ratio of 10% only applies to house­hold deposits; cor­po­rate deposits have no reserve require­ment. And the reserves are required with a 30 day lag after lend­ing has occurred—by which time the deposits cre­at­ed by the lend­ing are per­co­lat­ing through the bank­ing sys­tem.

Fig­ure 3: USA Reserve Require­ments

This, and the bank­ing cri­sis we are now in, final­ly inspired the Fed­er­al Reserve’s research depart­ment to con­clude that, effec­tive­ly, the “mon­ey mul­ti­pli­er” does­n’t exist. Car­pen­ter and Demi­ralp note that today reserve require­ments “are assessed on only about one-tenth of M₂″, and con­clude that

the nar­row, text­book mon­ey mul­ti­pli­er does not appear to be a use­ful means of assess­ing the impli­ca­tions of mon­e­tary pol­i­cy for future mon­ey growth or bank lend­ing. (Seth B. Car­pen­ter and Sel­va Demi­ralp, 2010, p. 29)

So Paul’s “gotcha” lacks at least one of the blades need­ed to make it work—and if he cares to con­sult the exten­sive aca­d­e­m­ic lit­er­a­ture on the role of reserves post the failed Mon­e­tarist exper­i­ment of the 1970s, he will see that the oth­er blade does­n’t exist either: Cen­tral Banks now sup­ply what­ev­er lev­el of reserves is need­ed to main­tain their short-run inter­est rate tar­get.

Who’s the Mystic then?

Krug­man’s claim that those who argue banks play an essen­tial role in macro­eco­nom­ics are “Bank­ing Mys­tics” has a nat­ur­al riposte: Neo­clas­si­cal econ­o­mists like Krug­man who believe that cap­i­tal­ism can be mod­elled with­out either mon­ey or banks are Barter Mys­tics (David Grae­ber, 2011). How on earth can some­one believe that the man­i­fest real­i­ty that trans­ac­tions involve mon­ey being exchanged for goods can be ignored, and pre­tend instead that goods are exchanged for goods? How on earth can the insti­tu­tion­al real­i­ty of banks be ignored by those who claim to be macro­econ­o­mists?

How on earth indeed. It’s because they’re still liv­ing in a pre-Coper­ni­can uni­verse, delud­ed by the imag­ined per­fec­tion of the Spheres.

Cardall. 2000. “The Uni­verse of Aris­to­tle and Ptole­my,”

Car­pen­ter, Seth B. and Sel­va Demi­ralp. 2010. “Mon­ey, Reserves, and the Trans­mis­sion of Mon­e­tary Pol­i­cy: Does the Mon­ey Mul­ti­pli­er Exist?,” Finance and Eco­nom­ics Dis­cus­sion Series. Wash­ing­ton: Fed­er­al Reserve Board,

Grae­ber, David. 2011. Debt: The First 5,000 Years. New York: Melville House.

Jor­da, Oscar; Moritz H. P. Schu­lar­ick and Alan M. Tay­lor. 2011a. “When Cred­it Bites Back: Lever­age, Busi­ness Cycles, and Crises,” Nation­al Bureau of Eco­nom­ic Research, Inc, NBER Work­ing Papers: 17621,

Jor­da, Oscar; Moritz Schu­lar­ick and Alan M. Tay­lor. 2011b. “When Cred­it Bites Back: Lever­age, Busi­ness Cycles, and Crises,” Fed­er­al Reserve Bank of San Fran­cis­co, Work­ing Paper Series: 2011–27,

O’Brien, Yueh-Yun June C. 2007. “Reserve Require­ment Sys­tems in Oecd Coun­tries.” SSRN eLi­brary.