Debtwatch No. 38: The GFC—Pothole or Mountain?

Flattr this!

The Marx­i­an view is that cap­i­tal­is­tic economies are inher­ent­ly unsta­ble and that exces­sive accu­mu­la­tion of cap­i­tal will lead to increas­ing­ly severe eco­nom­ic crises. Growth the­o­ry, which has proved to be empir­i­cal­ly suc­cess­ful, says this is not true.

The cap­i­tal­is­tic econ­o­my is sta­ble, and absent some change in tech­nol­o­gy or the rules of the eco­nom­ic game, the econ­o­my con­verges to a con­stant growth path with the stan­dard of liv­ing dou­bling every 40 years.

In the 1930s, there was an impor­tant change in the rules of the eco­nom­ic game. This change low­ered the steady-state mar­ket hours. The Key­ne­sians had it all wrong.

In the Great Depres­sion, employ­ment was not low because invest­ment was low. Employ­ment and invest­ment were low because labor mar­ket insti­tu­tions and indus­tri­al poli­cies changed in a way that low­ered nor­mal employ­ment.”

Obvi­ous­ly, I did not write the above. The author was instead Edward C. Prescott, who shared the 2004 Nobel Prize in Eco­nom­ics for the devel­op­ment of real busi­ness cycle the­o­ry, in his 1999 paper “Some Obser­va­tions on the Great Depres­sion” (Fed­er­al Reserve Bank of Min­neapo­lis Quar­ter­ly Review, Win­ter 1999, vol. 23, no. 1, pp. 25– 31).

This state­ment is remark­able for a num­ber of rea­sons I’ll dis­cuss below. But though it is extreme, it does express a belief that is endem­ic in neo­clas­si­cal eco­nom­ics, that a mar­ket econ­o­my is inher­ent­ly sta­ble and will always return to a sta­ble growth path after a shock.

That com­mon belief lies behind the expec­ta­tions of econ­o­mists that, now that the GFC has played itself out, the econ­o­my will return to trend growth and the emer­gency mea­sures that atten­u­at­ed its impact can be with­drawn.

From this per­spec­tive, the GFC was a “pot­hole in the road” caused by the Sub­prime cri­sis, a “change in the rules of the eco­nom­ic game” which is now behind us. With the dam­age caused by the cri­sis large­ly con­tained, nor­mal eco­nom­ic growth can resume. Over time, the unem­ploy­ment rate will return to pre-cri­sis lev­els as the eco­nom­ic car resumes its steady speed along the high­way of his­to­ry.

The alter­na­tive per­spec­tive is that the GFC was more akin to an abrupt change in the ter­rain. The “eco­nom­ic car” had been coast­ing down­hill with the grav­i­ty of ever-increas­ing pri­vate debt adding to the speed of the car. With the GFC we reached the bot­tom of the hill, and the car now has to dri­ve uphill as it attempts to main­tain its pre­vi­ous debt-enhanced speed while also reduc­ing debt.

Visu­al­ly at least, the “change in ter­rain” anal­o­gy stands up bet­ter than the pot­hole. I nor­mal­ly show the debt to GDP ratio as a ris­ing func­tion, but the econ­o­my’s speed gets a boost as the increase in debt makes a pos­i­tive con­tri­bu­tion to aggre­gate demand, and is slowed down when delever­ag­ing reduces demand. So turn­ing the ratio upside down may give a bet­ter idea of the depth of the “Val­ley of Debt” into which we have fall­en:

When Aus­tralia began its most recent descent into debt in mid-1964, the aver­age annu­al increase of 4.2% in the ratio added only a triv­ial amount to aggre­gate demand—since at the time debt was a mere 25% of GDP. But at the end of the debt bub­ble in 2008, when debt had become 165% of GDP, that same rate of debt growth added a huge amount to demand—the eco­nom­ic “car” gained speed as the slope of the debt moun­tain increased.

We hit the bot­tom of that moun­tain in March 2008, and now we’re start­ing to climb out of the valley—though not yet in absolute terms, since thanks to the First Home Ven­dors Boost, mort­gage debt is still grow­ing as busi­ness busi­ly delevers (see com­ments on the data, below). But once delever­ag­ing takes hold, the accel­er­a­tion caused by rac­ing down Debt Moun­tain will be replaced by an eco­nom­ic car strain­ing up the Mount Debt Reduc­tion. This change in the ter­rain will con­strain pri­vate eco­nom­ic per­for­mance until debt has fall­en sig­nif­i­cant­ly, as it did after the 1890s and the 1930s.

A sim­i­lar, if more extreme, pic­ture applies in the USA, where pri­vate debt is now 300% of GDP. In con­trast to Aus­tralia, the USA’s debt ratio began to rise as soon as WWII end­ed: on aver­age, US pri­vate debt rose 2.9% faster than GDP every year until 2008, tak­ing the debt ratio from 45% at the end of the War to 300% now. Delever­ag­ing from this lev­el of debt must exert a sub­stan­tial break on eco­nom­ic per­for­mance, by divert­ing income from expen­di­ture to debt reduc­tion.

I am there­fore one of a minor­i­ty of eco­nom­ic com­men­ta­tors who regard “defla­tion and delever­ag­ing” as the main dan­gers fac­ing the glob­al econ­o­my in the near future (curi­ous­ly, this minor­i­ty might include Aus­tralian Prime Min­is­ter Kevin Rudd). From my per­spec­tive, the Glob­al Finan­cial Cri­sis marks “a change in the ter­rain”: for decades, ris­ing debt has tur­bocharged eco­nom­ic per­for­mance; now falling debt will be a drag on eco­nom­ic activ­i­ty.

The vast major­i­ty of econ­o­mists who per­ceive the GFC as a pot­hole on the road that is now behind us do not con­sid­er debt and delever­ag­ing in their analy­sis. Their mod­els have nei­ther cred­it nor mon­ey nor pri­vate debt in them, so from their point of view, there is no ter­rain at all beneath the car—merely a long flat high­way of his­to­ry along which the eco­nom­ic car dri­ves at the speed it is under­ly­ing “real” eco­nom­ic per­for­mance.

This fail­ure to even con­sid­er the role of pri­vate cred­it in a cap­i­tal­ist econ­o­my is an endem­ic weak­ness in con­ven­tion­al “neo­clas­si­cal” eco­nom­ics, which ignores the dynam­ics of cred­it for a vari­ety of rea­sons that are both ide­o­log­i­cal and illog­i­cal.

The ide­ol­o­gy is appar­ent in Prescot­t’s com­ments on the Great Depres­sion, quot­ed above. The lack of log­ic is evi­dent when you com­pare a key state­ment in that paper—that “Growth the­o­ry, which has proved to be empir­i­cal­ly suc­cess­ful, says this is not true”—with the results of some very care­ful empir­i­cal research by the very same author just ten years ear­li­er. There he (and co-author and Nobel Prize recip­i­ent Finn Kyd­land) con­clud­ed that the empir­i­cal data con­tra­dict­ed neo­clas­si­cal growth the­o­ry:

The pur­pose of this arti­cle is to present the busi­ness cycle facts in light of estab­lished neo­clas­si­cal growth the­o­ry, which we use as the orga­niz­ing frame­work for our pre­sen­ta­tion of busi­ness cycle facts. We empha­size that the sta­tis­tics report­ed here are not mea­sures of any­thing; rather, they are sta­tis­tics that dis­play inter­est­ing pat­terns, giv­en the estab­lished neo­clas­si­cal growth the­o­ry.

In dis­cus­sions of busi­ness cycle mod­els, a nat­ur­al ques­tion is, Do the cor­re­spond­ing sta­tis­tics for the mod­el econ­o­my dis­play these pat­terns? We find these fea­tures inter­est­ing because the pat­terns they seem to dis­play are incon­sis­tent with the the­o­ry.” (Finn E. Kyd­land & Edward C. Prescott, “Busi­ness Cycles: Real Facts and a Mon­e­tary Myth”, Fed­er­al Reserve Bank of Min­neapo­lis Quar­ter­ly Review, vol. 14, no. 2, pp 3–18, p. 4).

One key pat­tern in actu­al eco­nom­ic data that went against the pre­dic­tions of neo­clas­si­cal eco­nom­ic the­o­ry was the rela­tion­ship between broad mea­sures of the mon­ey sup­ply and gov­ern­ment-cre­at­ed “Base Mon­ey”. The stan­dard “mon­ey mul­ti­pli­er” view is that:

  1. The gov­ern­ment cre­ates “Base Mon­ey” via deficit spend­ing, and cred­its that mon­ey to pri­vate indi­vid­u­als via social secu­ri­ty, goods pur­chas­es etc.;
  2. These pri­vate indi­vid­u­als then deposit that mon­ey in bank accounts;
  3. The banks then retain a pro­por­tion of these deposits and lend out the rest, cre­at­ing cred­it mon­ey (and debt).

If this view were empir­i­cal­ly cor­rect, then an analy­sis of mon­ey over time would show that “Base Mon­ey” was cre­at­ed first and “Cred­it Mon­ey” was cre­at­ed lat­er, with a time lag.

In fact, what Kyd­land and Prescott found was that the empir­i­cal data was the oppo­site of this: cred­it mon­ey was cre­at­ed first, and Base Mon­ey was cre­at­ed lat­er, with a lag of up to a year:

There is no evi­dence that either the mon­e­tary base or M1 leads the cycle, although some econ­o­mists still believe this mon­e­tary myth. Both the mon­e­tary base and M 1 series are gen­er­al­ly pro­cycli­cal and, if any­thing, the mon­e­tary base lags the cycle slight­ly.

The dif­fer­ence in the behav­ior of M1 and M2 sug­gests that the dif­fer­ence of these aggre­gates (M2 minus M1) should be con­sid­ered. … The dif­fer­ence of M2-M1 leads the cycle by even more than M2, with the lead being about three quar­ters.

The fact that the trans­ac­tion com­po­nent of real cash bal­ances (M 1) moves con­tem­po­ra­ne­ous­ly with the cycle while the much larg­er non­trans­ac­tion com­po­nent (M2) leads the cycle sug­gests that cred­it arrange­ments could play a sig­nif­i­cant role in future busi­ness cycle the­o­ry. Intro­duc­ing mon­ey and cred­it into growth the­o­ry in a way that accounts for the cycli­cal behav­ior of mon­e­tary as well as real aggre­gates is an impor­tant open prob­lem in eco­nom­ics.

I could­n’t agree more, but this is not what neo­clas­si­cal econ­o­mists did. Instead they con­tin­ued to devel­op mod­els in which mon­ey and debt played no role.

Despite his excel­lent empir­i­cal work on mon­e­tary dynam­ics in “Real Facts and a Mon­e­tary Myth”, Prescot­t’s “Great Depres­sion” paper made no ref­er­ence to cred­it at all as an explana­to­ry fac­tor in the Great Depres­sion. Instead—I’m not joking—he blamed the Depres­sion on a “change in labor mar­ket insti­tu­tions and indus­tri­al poli­cies that low­ered steady-state, or nor­mal, mar­ket hours”.

Except for this bizarre argu­ment that the Great Depres­sion was the result of the vol­un­tary response of work­ers to unspec­i­fied changes in labour mar­ket con­di­tions that made labour less desir­able, this lengthy quote from Prescott is rep­re­sen­ta­tive of stan­dard neo­clas­si­cal think­ing about crises like the GFC:

Essen­tial­ly, busi­ness cycles are respons­es to per­sis­tent changes, or shocks, that shift the con­stant growth path of the econ­o­my up or down. This con­stant growth path is the path to which the econ­o­my would con­verge if there were no sub­se­quent shocks. If a shock shifts the con­stant growth path down, the econ­o­my responds as fol­lows. Mar­ket hours fall, reduc­ing out­put; a big­ger share of out­put is allo­cat­ed to con­sump­tion and a small­er share to invest­ment; and more time is allo­cat­ed to leisure. Over time, mar­ket hours return to nor­mal, as do invest­ment and con­sump­tion shares of out­put, as the econ­o­my con­verges to its new low­er con­stant growth path. The lev­el of the new path is low­er, not the growth rate along the path.

I’ve just described the response of the econ­o­my to a sin­gle shock. In fact, the econ­o­my is con­tin­u­al­ly hit by shocks, and what econ­o­mists observe in busi­ness cycles is the effects of past and cur­rent shocks. A bust occurs if a num­ber of neg­a­tive shocks are bunched in time. A boom occurs if a num­ber of pos­i­tive shocks are bunched in time. Busi­ness cycles are, in the lan­guage of Slutzky (1937), the “sum of ran­dom caus­es.”

The fun­da­men­tal dif­fer­ence between the Great Depres­sion and busi­ness cycles is that mar­ket hours did not return to nor­mal dur­ing the Great Depres­sion. Rather, mar­ket hours fell and stayed low. In the 1930s, labor mar­ket insti­tu­tions and indus­tri­al pol­i­cy actions changed nor­mal mar­ket hours. I think these insti­tu­tions and actions are what caused the Great Depres­sion.”

So the Great Depres­sion was a con­scious choice by Amer­i­can work­ers to enjoy more leisure, in response to unspec­i­fied changes in the labour mar­ket  ([Lat­er in the same essay, he states: “Exact­ly what changes in mar­ket insti­tu­tions and indus­tri­al poli­cies gave rise to the large decline in nor­mal mar­ket hours is not clear.…”).

It would be bad enough if Prescott were mere­ly an obscure aca­d­e­m­ic econ­o­mist, but he is far from obscure: he and Kyd­land shared the Nobel Prize in Eco­nom­ics for the devel­op­ment of neo­clas­si­cal growth the­o­ry. As ridicu­lous as his argu­ment is, it does accu­rate­ly state the con­clu­sions of the neo­clas­si­cal “real busi­ness cycle” mod­el. As is often the case, you find a much clearer—and there­fore far more obvi­ous­ly absurd—statement of neo­clas­si­cal eco­nom­ic the­o­ry when you go to the source, rather than rely­ing on a sec­ond-hand account from a text­book or run-of-the-mill prac­ti­tion­er.

So the con­fi­dence that the vast major­i­ty of econ­o­mists have that the GFC is now behind us, and the “nor­mal” trend rate of growth will resume, is fun­da­men­tal­ly based on the belief that cred­it and debt dynam­ics do not mat­ter.

I beg to dif­fer. Though the enor­mous gov­ern­ment stim­u­lus has atten­u­at­ed the imme­di­ate impact of debt delever­ag­ing, it has done noth­ing to reduce the out­stand­ing lev­el of pri­vate debt. Instead even sub-par growth has become depen­dent on con­tin­u­ing gov­ern­ment stim­uli, and when­ev­er those stim­uli are removed, the econ­o­my will fal­ter.

Total pri­vate debt rose by a mere A$1 bil­lion last month, ver­sus as much as A$30 bil­lion dur­ing the height of the debt bub­ble. But were it not for the First Home Ven­dors Boost (let’s call it what it is), Aus­tralia would now be firm­ly in the grips of delever­ag­ing.


COMMENTS ON THE DATA—A Mortgage & Government Led Recovery?

Total pri­vate debt rose by a mere A$1 bil­lion last month, ver­sus as much as A$30 bil­lion dur­ing the height of the debt bub­ble. But were it not for the First Home Ven­dors Boost (let’s call it what it is), Aus­tralia would now be firm­ly in the grips of delever­ag­ing.

Nonethe­less the debt to GDP ratio fell yet again, because the rate of growth of debt is now sub­stan­tial­ly below the rate of growth of GDP—even though that is now also anaemic.

The break­down of debt shows that the busi­ness sec­tor is rapid­ly delever­ag­ing, while mort­gage and gov­ern­ment debt is escalating—and both those are the result of gov­ern­ment pol­i­cy.

With­out the First Home Ven­dors Boost, it is high­ly unlike­ly that mort­gage debt would still be ris­ing today. Mort­gage debt peaked as a per­cent­age of GDP in March 2008, and fell for the remain­der of the year until the First Home Ven­dors Boost.

The quar­ter­ly change in mort­gage debt was also trend­ing down from the 2005 peak, and that down­ward trend has clear­ly been reversed by the impact of the Boost.

House Prices

The Boost has cer­tain­ly had the impact the gov­ern­ment desired, of arrest­ing the fall in Aus­tralian house prices.

It will also almost cer­tain­ly guar­an­tee that I’ll be walk­ing (and run­ning) to Kosciuszko under the first half of the bet with Rory Robert­son.[1] The sec­ond half of the bet, that the fall from peak to trough will be of the order of 40%, may still see Rory also walk­ing some years hence—and the with­draw­al of the Boost may make this occur soon­er rather than lat­er.

The rea­son is twofold. First­ly, the Boost has obvi­ous­ly brought for­ward some buy­ing by First Home Buy­ers that would have occurred any­way, as well as entic­ing in oth­ers who might not have con­sid­ered it oth­er­wise. The with­draw­al of that demand will have a strong impact on the sub-$500,000 price range.

But the with­draw­al will also affect hous­es in the $1 mil­lion to $1.5 mil­lion range as well, because the Boost did far more than mere­ly boost sub-$500,000 prices.

First Home Buy­ers who were enticed into the mar­ket by the addi­tion­al $7,000 geared that up with addi­tion­al debt by at least a fac­tor of 4, to result in some­thing like a  $35,000 price jump for sub-$500K hous­es. But the sell­ers of those houses—the real ben­e­fi­cia­ries of the Boost—then received an extra $35,000 in cold hard cash. They then used this as a boost to their own deposits on their pur­chas­es of hous­es fur­ther up the chain—and if they also geared by a fac­tor of 4 (ie a 80% mar­gin­al lev­el of gear­ing, which is well with­in cur­rent lend­ing prac­tice), then the prices they paid for hous­es in the $750K‑1.5M range would have risen by $140,000.

This works in reverse as well. When the Boost is with­drawn, not only will sell­ers of sub-$500K hous­es find that buy­ers have $35K less to spend than dur­ing the boost, the sell­ers of $750K‑1.5M abodes will find their buy­ers short about $150K com­pared to dur­ing the boost.

2010 could be an inter­est­ing year for Aus­tralian house prices.

[1] If the index breaks its cur­rent max­i­mum lev­el of 131 in the next release of ABS 6416, I will walk (and run) from Par­lia­ment House to Mt Kosciusko as required by the bet in the last weeks of Feb­ru­ary 2010.

Bookmark the permalink.