DebtWatch No 28 November 2008: What is Really Going On?

Flattr this!

2nd Anniver­sary Issue…

Why Did I See it Coming and “They” Didn’t?

The finan­cial cri­sis is wide­ly accept­ed as hav­ing start­ed in August 9 2007, with the BNP’s announce­ment that it was sus­pend­ing redemp­tions from three of its funds that were heav­i­ly exposed to the US secu­ri­ti­sa­tion mar­ket (click here for the BNP August 9 2007 press release).

Just three months before­hand, the OECD released its 2007 World Eco­nom­ic Out­look, in which it com­ment­ed that:

In its Eco­nom­ic Out­look last Autumn, the OECD took the view that the US slow­down was not herald­ing a peri­od of world­wide eco­nom­ic weak­ness, unlike, for instance, in 2001. Rather, a “ smooth”  rebal­anc­ing was to be expect­ed, with Europe tak­ing over the baton from the Unit­ed States in dri­ving OECD growth.

Recent devel­op­ments have broad­ly con­firmed this prog­no­sis. Indeed, the cur­rent eco­nom­ic sit­u­a­tion is in many ways bet­ter than what we have expe­ri­enced in years. Against that back­ground, we have stuck to the rebal­anc­ing sce­nario. Our cen­tral fore­cast remains indeed quite benign: a soft land­ing in the Unit­ed States, a strong and sus­tained recov­ery in Europe, a sol­id tra­jec­to­ry in Japan and buoy­ant activ­i­ty in Chi­na and India. In line with recent trends, sus­tained growth in OECD economies would be under­pinned by strong job cre­ation and falling unem­ploy­ment.” (OECD World Eco­nom­ic Out­look Vol 81 p. 7; emphases added)

Sim­i­lar­ly, Reserve Banks around the world had set inter­est rates to rel­a­tive­ly high lev­els to restrain ris­ing infla­tion, which was then seen as the main threat to con­tin­ued eco­nom­ic pros­per­i­ty. Our own RBA increased rates when the cri­sis began, and three more times since. And it was not alone: the Euro­pean Cen­tral Bank also raised rates after the cri­sis (see Fig­ure One) .

Figure One

Reserve Interest Rates

Reserve Inter­est Rates

In Decem­ber 2005, almost two years before the cri­sis hit, I realised that a seri­ous finan­cial cri­sis was approach­ing. I was so wor­ried about its prob­a­ble severity–and the lack of aware­ness about it amongst pol­i­cy makers–that I took the risk (for an aca­d­e­m­ic) of going very pub­lic about my views. I began com­ment­ing on eco­nom­ic pol­i­cy in the media, start­ed the Debt­Watch Report (the first was pub­lished two years ago in Novem­ber 2006), reg­is­tered a web­page with the apt name of, and estab­lished the blog Steve Keen’s Oz Debt­watch.

How come I got it right, and “they”–the offi­cial eco­nom­ic managers–got it so wrong?

It’s not because I’m any brighter than they are–there are plen­ty of high­ly intel­li­gent peo­ple in those organ­i­sa­tions. Instead, it’s because they fol­low main­stream views in eco­nom­ics, and I fol­low a minor­i­ty per­spec­tive. The eco­nom­ic his­to­ry we are cur­rent­ly liv­ing through is proof that the main­stream is fun­da­men­tal­ly wrong about the nature of the econ­o­my, while my minor­i­ty per­spec­tive is at least par­tial­ly right.

This is not some­thing one should be able to say about a sci­ence, and there lies the rub: eco­nom­ics is not even close to qual­i­fy­ing as a sci­ence. A bet­ter mod­el for eco­nom­ics is a group of war­ring religions–or sci­ence, such as it was, before Galileo’s empir­i­cal rev­o­lu­tion, when what mat­tered to sci­en­tists was not empir­i­cal rel­e­vance, but con­for­mi­ty to with the Bible.

Forty years ago, Keynes was The Mes­si­ah, and his Gen­er­al The­o­ry was the Bible. But the “stagfla­tion” episode of the 1970s allowed a new Mes­si­ah to arise: Mil­ton Fried­man, with his doc­trine of Mon­e­tarism. Though Mon­e­tarism itself is no longer espoused, the eco­nom­ic reli­gion that Fried­man represented–known as “Neo­clas­si­cal Economics”–supplanted the pre­vi­ous Key­ne­sian ortho­doxy. Today, the major­i­ty of econ­o­mists know of no oth­er way to think about the economy–and they run Cen­tral Banks and Trea­suries through­out the world, and dom­i­nate tuition in uni­ver­si­ties.

They also devel­op math­e­mat­i­cal mod­els of the econ­o­my, which are in turn used to guage its health, and to advise politi­cians about pol­i­cy chal­lenges in the near future. Accord­ing to these mod­els, just over a year ago the econ­o­my was in fine shape, and the main pol­i­cy chal­lenge was to avoid over­heat­ing that would lead to ris­ing infla­tion.

Well infla­tion did rise. But simul­ta­ne­ous­ly the glob­al econ­o­my was falling into a seri­ous reces­sion dri­ven by a glob­al finan­cial melt­down that these econ­o­mists and their mod­els com­plete­ly failed to antic­i­pate. Rarely in human his­to­ry have pol­i­cy mak­ers been so bad­ly mis­led by the so-called experts.

The three key aspects of Neo­clas­si­cal eco­nom­ics that led to its wild­ly inac­cu­rate fore­casts are the beliefs that:

  1. A mar­ket econ­o­my always tends towards equi­lib­ri­um;
  2. Mon­ey impacts “nom­i­nal” vari­ables like the rate of infla­tion, but has no long term impact on “real” vari­ables like employ­ment and GDP growth; and
  3. Finance mar­kets are ratio­nal; in par­tic­u­lar, the lev­el of pri­vate debt reflects ratio­nal cal­cu­la­tions about future income, and can there­fore be ignored by pol­i­cy-mak­ers.

The key aspects of the approach that I take (the “Finan­cial Insta­bil­i­ty Hypoth­e­sis” devel­oped by Hyman Min­sky) that alert­ed me to the approach­ing dan­ger are the propo­si­tions that:

  1. A mar­ket econ­o­my is inher­ent­ly cycli­cal;
  2. Mon­ey is fun­da­men­tal­ly cred­it-dri­ven, and has impacts on real vari­ables as well as nom­i­nal ones in the short and long term; and
  3. Finance mar­kets desta­bilise the real econ­o­my, because they are prone to bouts of euphor­ic expec­ta­tions that lead to debt-financed spec­u­la­tive bub­bles.

These very dif­fer­ent per­spec­tives have two key effects on the econ­o­mists who hold them:

  • they focus atten­tion on very dif­fer­ent sets of eco­nom­ic data; and
  • they inspire math­e­mat­i­cal mod­els of the econ­o­my that are very, very dif­fer­ent.

What is “a beautiful set of numbers” lies in the eyes of the beholder

Neo­clas­si­cal econ­o­mists focus upon three num­bers:

  1. The rate of eco­nom­ic growth (prefer­ably above 3% per year);
  2. Unem­ploy­ment (which they pre­fer to be low, but not “too low”–the mov­ing tar­get for which in Aus­tralia was 4.5% until recent­ly); and
  3. The rate of infla­tion (which they pre­fer to be as low as pos­si­ble, and cer­tain­ly below 3%).

On all three fronts, from the van­tage point of 2006, 2007 looked like being a vin­tage year–except that the first num­ber was so high that the sec­ond was tend­ing too low, which could mean that the third could start to rise. Hence the eco­nom­ic focus was on reduc­ing growth via high­er inter­est rates, to increase unem­ploy­ment slight­ly and thus reduce the rate of infla­tion (see Fig­ure Two).

Figure Two

Australian Growth, Unemployment and Inflation

Aus­tralian Growth, Unem­ploy­ment and Infla­tion

The RBA’s pol­i­cy response to this was imme­di­ate and deci­sive. Hav­ing already raised rates in 0.25% incre­ments five times since 2002, it accel­er­at­ed its infla­tion-con­trol pro­gram with three more increas­es in 2006, two in 2007 –the first of these com­ing just before the cri­sis broke, and the oth­er famous­ly dur­ing the 2007 elec­tion campaign–and two more dur­ing ear­ly 2008.

Figure Three

Movements in Australian Reserve Interest Rates

Move­ments in Aus­tralian Reserve Inter­est Rates

Unfor­tu­nate­ly, the RBA’s response was also the wrong one. While infla­tion did rise, it was not the main prob­lem fac­ing the econ­o­my. Try­ing to con­trol the infla­tion rate by rais­ing inter­est rates at that time was a bit like try­ing to con­trol a patien­t’s blood pres­sure when he was dying of can­cer. That can­cer, as is now wide­ly acknowl­edged, was pri­vate debt. The eco­nom­ic vari­able that their Neo­clas­si­cal train­ing led them to ignore, the ratio of pri­vate debt to GDP, was now indis­putably the most impor­tant num­ber of all.

Econ­o­mists who are influ­enced by Hyman Minsky–broadly known as “Post Key­ne­sians”, since Min­sky was a fol­low­er of Keynes–focus pre­cise­ly on that datum. This ratio of a stock (out­stand­ing debt at a point in time) to a flow (the annu­al out­put of goods and ser­vices) tells you how many years of income it would take to reduce debt to zero. It there­fore mea­sures the degree of pres­sure that finance is impos­ing on the real econ­o­my.

A cer­tain amount of debt is vital to the prop­er func­tion­ing of a mar­ket econ­o­my, since most com­pa­nies need flex­i­ble work­ing cap­i­tal to be able to oper­ate, and over­draft facil­i­ties and lines of cred­it pro­vide that flex­i­bil­i­ty. But too high a debt to GDP ratio means that the finan­cial bur­den of debt repay­ment on the econ­o­my is exces­sive, and Min­sky’s the­o­ry implies that there is a ten­den­cy for the debt to GDP ratio to ratch­et up over a series of booms and busts, result­ing even­tu­al­ly to a Depres­sion.

I did not see the data in Fig­ure Three until Decem­ber 2005, since my “day job” is as an aca­d­e­m­ic rather than an eco­nom­ic pol­i­cy mak­er. I had signed a con­tract to pro­duce a book on finan­cial insta­bil­i­ty as long ago as 1998, but the unex­pect­ed suc­cess of Debunk­ing Eco­nom­ics, and the fol­low-on debate that engen­dered amongst aca­d­e­m­ic econ­o­mists, forced me to delay com­menc­ing that task.

As soon as I did see the data–in Decem­ber 2005, when prepar­ing an Expert Wit­ness report for a court case (the “Cooks Case”)–my Min­skian eyes told me that a seri­ous cri­sis was on its way. Giv­en that the debt to GDP ratio was far high­er than dur­ing either major post-WWII cri­sis (1973 and 1989), it appeared obvi­ous that Aus­tralia was about to expe­ri­ence its most severe eco­nom­ic cri­sis since the Great Depres­sion.

Because I knew that neo­clas­si­cal econ­o­mists would not realise this was about to hap­pen, were like­ly to make things worse by increas­ing inter­est rates as the cri­sis approached, and would prob­a­bly mis-diag­nose the cause once it occurred–as they had dur­ing the Great Depression–I decid­ed to go pub­lic with my analy­sis via the media, a reg­u­lar com­men­tary timed to coin­cide with the RBA meet­ing (Debt­Watch), and even­tu­al­ly a blog (

Figure Four

Australias Debt to GDP Ratio 1955-Now

Aus­trali­a’s Debt to GDP Ratio 1955-Now

Min­sky’s hypoth­e­sis warns that a cri­sis begins with the fal­ter­ing of an asset price bub­ble. That not one but two bub­bles were in progress was obvi­ous in both Aus­tralian and Amer­i­can stock mar­ket and hous­ing data.

Min­sky argues that there are two price lev­els in a mar­ket economy–one for com­modi­ties set large­ly by the costs of pro­duc­tion and financed large­ly from income, and the oth­er for assets, set large­ly by peo­ple’s expec­ta­tions of future gain, and financed main­ly by debt. The ratio of one price lev­el to anoth­er thus gives an indi­ca­tion of whether the econ­o­my is in a bub­ble, or a bust.

There are curly issues in the ratio of share prices to the CPI–the rein­vest­ment of retained earn­ings give shares an upward trend over time com­pared to the CPI, while the index itself over­states share returns due to sur­vivor bias. But the rel­a­tive­ly rapid move­ment in share prices, ver­sus the slow­er changes in con­sumer prices, means that a blowout in the ratio is a good indi­ca­tor of a bub­ble. On that basis, Aus­trali­a’s mar­ket had clear­ly entered a bub­ble in ear­ly 2003, while the USA’s began in 1995 (and had already burst in 2000, only to restart in 2003).

Figure Five

CPI-Deflated Stock Market Indices, USA and Australia

No such curly issues apply with the house price to CPI ratio. Espe­cial­ly when deal­ing with estab­lished hous­es, there is no long term trend, as the Heren­gracht Canal index estab­lish­es. In a real price series going back over 300 years, house prices have risen and fall­en com­pared to con­sumer prices, but there is clear­ly no ris­ing trend (see Fig­ure Five).

Figure Six

CPI-Deflated Price Index for Amsterdams Herengracht Canal

CPI-Deflat­ed Price Index for Ams­ter­dam’s Heren­gracht Canal

On this basis, both Aus­tralian and US house mar­kets were clear­ly in bub­bles, and the US bub­ble was unprece­dent­ed in its his­to­ry.

Figure Seven

CPI-Deflated Herengracht Canal Index: 350 Years from 1928-1970

USA and Aus­tralian House Price Indices

The final piece of evi­dence that pushed me from expect­ing a seri­ous reces­sion to quite pos­si­bly a Depres­sion was pro­vid­ed by the RBA in Sep­tem­ber 2007–a month after the cri­sis began–with a chart show­ing Aus­trali­a’s pri­vate debt to GDP ratio going back till 1860.

Even after I aug­ment­ed it to include an esti­mate for non-bank debt pri­or to 1953 (which made cur­rent data look less extreme com­pared to his­tor­i­cal data), it implied that our debt cri­sis was more than twice as severe as the one that caused the Great Depres­sion. When the Great Depres­sion began at the end of 1929, Aus­trali­a’s debt to GDP ratio was 65 per­cent. It has now reached a peak of 165 per­cent.

Figure Eight

Australias Debt to GDP Ratio from 1860-Now

Aus­trali­a’s Debt to GDP Ratio from 1860-Now

That impres­sion was con­firmed when I lat­er saw the US data–courtesy of Ger­ard Minack and the avail­abil­i­ty online of US Cen­sus reports. Its debt to GDP ratio was 150 per­cent at the end of 1929 (and sub­se­quent­ly blew out to 215 per­cent as prices and GDP col­lapsed in the first 3 years of the Depres­sion).

With finan­cial sec­tor debt includ­ed, the USA reached that peak again in 1987–the year Greenspan, despite his “Aus­tri­an” approach to eco­nom­ics that decried gov­ern­ment inter­ven­tion of any sort, per­formed his first “suc­cess­ful” res­cue dur­ing the Stock Mar­ket Crash in Octo­ber.

That res­cue worked, not by over­com­ing the prob­lem of exces­sive debt-financed spec­u­la­tion, but by re-ignit­ing so that it reached even high­er lev­els. Though bor­row­ing slumped after the Sav­ings and Loans col­lapse in 1989/90–falling from 170 to 165 per­cent of GDP–the bub­ble began once more in 1994. It then rock­et­ed on through the Dot­com Bub­ble, and did­n’t even draw breath then since there were now two asset mar­ket bub­bles feed­ing off each other–the Sub­prime Bub­ble’s expan­sion more than coun­ter­act­ed the Dot­com Bub­ble’s coll­pase, until final­ly there were two debt-financed asset bub­bles run­ning at once–an unprece­dent­ed event in Amer­i­ca’s finan­cial his­to­ry.

By 2004, even non-finan­cial pri­vate debt had exceed­ed the lev­el that trig­gered the Great Depres­sion, while total pri­vate sec­tor debt reached a stag­ger­ing 290 per­cent of GDP (with­out includ­ing the impact of finan­cial deriv­a­tives, anoth­er form of debt that did not exist in the 1920s).

Figure Nine

USAs Long Term Debt to GDP Ratio 1920-Now

USA’s Long Term Debt to GDP Ratio 1920-Now

The RBA data in Fig­ure Eight was first pub­lished in a speech by Deputy RBA Gov­er­nor Ric Bat­telli­no (“Some Obser­va­tions on Finan­cial Trends”). I found his inter­pre­ta­tion of the chart both stun­ning, and pre­dictable:

The fac­tors that have facil­i­tat­ed the rise in debt over the past cou­ple of decades –  the sta­bil­i­ty in eco­nom­ic con­di­tions and the con­tin­ued flow of inno­va­tions com­ing from a com­pet­i­tive and dynam­ic finan­cial sys­tem –  remain in place. While ever this is the case, house­holds are like­ly to con­tin­ue to take advan­tage of unused capac­i­ty to increase debt. This is not to say that there won’ t be cycles when cred­it grows slow­ly for a time, or even falls, but these cycles are like­ly to take place around a ris­ing trend. Even­tu­al­ly, house­hold debt will reach a point where it is in some form of equi­lib­ri­um rel­a­tive to GDP or income, but the evi­dence sug­gests that this point is high­er than cur­rent lev­els.” (empha­sis added)

This was stun­ning, because the pre­vi­ous two peaks in the debt to GDP ratio were fol­lowed by Depres­sions, and yet they were far low­er than the cur­rent lev­el. Even the most anec­do­tal approach to his­to­ry would imply that all might not be well at present.

It was pre­dictable, because it was con­sis­tent with the mind­set that has dom­i­nat­ed eco­nom­ics for three decades now, ever since Fried­man’s counter-rev­o­lu­tion against Key­ne­sian eco­nom­ics in the 1970s. Where­as the once-dom­i­nant Key­ne­sian approach saw the econ­o­my as poten­tial­ly unsta­ble, Fried­man’s revived “Neo­clas­si­cal” approach pre­sumed that the econ­o­my was self-equi­li­brat­ing. Thus data which an engi­neer would see as indi­cat­ing an approach­ing break­down was inter­pret­ed by an econ­o­mist as indi­cat­ing an approach­ing equi­lib­ri­um. 

This belief in a ten­den­cy to equi­lib­ri­um is built into the mod­els of the econ­o­my that neo­clas­si­cal econ­o­mists construct–which is why these mod­els gave no warn­ing of the approach­ing cri­sis, and why econ­o­mists were the last ones to realise that a cri­sis was actu­al­ly hap­pen­ing. I’ll dis­cuss their models–and the Min­skian alternative–in next mon­th’s Debt­watch.

Australian Private Aggregate Debt Table

Aus­tralian Pri­vate Aggre­gate Debt Table

Australias Private Debt Table Disaggregated

Aus­trali­a’s Pri­vate Debt Table Dis­ag­gre­gat­ed

Bookmark the permalink.

About Steve Keen

I am Professor of Economics and Head of Economics, History and Politics at Kingston University London, and a long time critic of conventional economic thought. As well as attacking mainstream thought in Debunking Economics, I am also developing an alternative dynamic approach to economic modelling. The key issue I am tackling here is the prospect for a debt-deflation on the back of the enormous private debts accumulated globally, and our very low rate of inflation.