It’s Hard Being a Bear (Part Five): Rescued?

Sep19
by Cassander on September 19, 2009 at 2:30 pm

flattr this!

I’m happy to admit that I under­es­ti­mated how strongly gov­ern­ments would respond to this finan­cial cri­sis. Dra­matic reduc­tions in inter­est rates, huge fis­cal stim­uli and—in the USA and UK—expansion of government-created money, have all had a pos­i­tive impact on the econ­omy and asset mar­kets (both shares and houses).

In his recent essay, Aus­tralian Prime Min­is­ter Kevin Rudd esti­mated that the res­cues were the equiv­a­lent of roughly 18 per­cent of global GDP over a 3 year period, which is an unprece­dented level of expen­di­ture by governments.

Eichen­green and O’Rourke’s com­par­i­son of today to the Great Depres­sion gives the most bal­anced assess­ment of how effec­tive these poli­cies have been at the global level.

They have clearly turned around stock mar­kets. Six months ago, world stock mar­kets were 50% below their peak, a far worse per­for­mance than dur­ing the Great Depres­sion when, at the same time after the peak, they had only fallen 10%. By the begin­ning of Sep­tem­ber, mar­kets had recov­ered to be only a cou­ple of per­cent below the com­pa­ra­ble 1930 posi­tion of a 30% fall.

Indus­trial out­put has also turned around. Six months ago this was 13% below the peak level, worse than the 1930s posi­tion of an 11% decline. Since then it has risen to be only 10% below, while at the equiv­a­lent time in the 1930s, indus­trial out­put had fallen 20% from its 1929 high.

So has the gov­ern­ment cav­alry rid­den to the res­cue? If the cri­sis were one sim­ply of liq­uid­ity, the answer would be yes. A gov­ern­ment stim­u­lus can over­whelm the impact of a credit crunch, and the innate dynamic of a pro­duc­tive econ­omy can re-assert itself after such a cri­sis, lead­ing to renewed growth.

But this not merely a cri­sis of liq­uid­ity. It is one of exces­sive pri­vate debt, on a scale that is also unprece­dented: the USA is car­ry­ing US$41.5 tril­lion in debt on the back of a US$14 tril­lion econ­omy, pro­por­tion­ately 70 per­cent more debt than it had at the start of the Great Depres­sion. In Decem­ber 2007, the pri­vate sec­tor swung from ramp­ing up debt lev­els as it chased spec­u­la­tive gains on asset mar­kets, to retreat­ing from debt as the asset bub­bles burst.

In the space of a year, pri­vate debt went from adding US$4 tril­lion to aggre­gate demand, to sub­tract­ing US$165 bil­lion from it. Pri­vate debt had ceased being the economy’s tur­bocharger and had instead become its flooded engine.

While eco­nomic out­siders like myself, Michael Hud­son, Niall Fer­gu­son and Nas­sim Taleb argue that the only way to restart the eco­nomic engine is to clear it of debt, the gov­ern­ment response, has been to attempt to replace the now defunct pri­vate debt eco­nomic tur­bocharger with a pub­lic one.

In the imme­di­ate term, the stu­pen­dous size of the stim­u­lus has worked, so that debt in total is still boost­ing aggre­gate demand. But what will hap­pen when the gov­ern­ment stops tur­bocharg­ing the econ­omy, and waits anx­iously for the pri­vate sys­tem to once again splut­ter into life?

I am afraid that all it will do is splutter.

This is espe­cially so since, fol­low­ing the advice of neo­clas­si­cal econ­o­mists, Obama has got not a bang but a whim­per out of the many bucks he has thrown at the finan­cial system.

In explain­ing his recov­ery pro­gram in April, Pres­i­dent Obama noted that:

“there are a lot of Amer­i­cans who under­stand­ably think that gov­ern­ment money would be bet­ter spent going directly to fam­i­lies and busi­nesses instead of banks – ‘where’s our bailout?,’ they ask”.

He jus­ti­fied giv­ing the money to the lenders, rather than to the debtors, on the basis of “the mul­ti­plier effect” from bank lending:

the truth is that a dol­lar of cap­i­tal in a bank can actu­ally result in eight or ten dol­lars of loans to fam­i­lies and busi­nesses, a mul­ti­plier effect that can ulti­mately lead to a faster pace of eco­nomic growth. (page 3 of the speech)

This argu­ment comes straight out of the neo­clas­si­cal eco­nom­ics text­book. For­tu­nately, due to the clear man­ner in which Obama enun­ci­ates it, the flaw in this text­book argu­ment is vividly appar­ent in his speech.

This “mul­ti­plier effect” will only work if Amer­i­can fam­i­lies and busi­nesses are will­ing to take on yet more debt: “a dol­lar of cap­i­tal in a bank can actu­ally result in eight or ten dol­lars of loans”.

So the only way the roughly US$1 tril­lion of money that the Fed­eral Reserve has injected into the banks will result in addi­tional spend­ing is if Amer­i­can fam­i­lies and busi­nesses take out another US$8–10 tril­lion in loans.

What are the odds that this will hap­pen, when they already owe more than they have ever owed in the his­tory of Amer­ica? The next chart inverts the usual por­trayal of America’s debt to GDP ratio by invert­ing it: the top of the graph rep­re­sents zero debt, the bot­tom, a debt to GDP ratio of 300 percent—which is just shy of the cur­rent ratio of 292 percent.

If the money mul­ti­plier was going to “ride to the res­cue”, pri­vate debt would need to rise from its cur­rent level of US$41.5 tril­lion to about US$50 tril­lion, and this ratio would rise to about 375%—more than twice the level that ush­ered in the Great Depression.

This is a res­cue? It’s a “hair of the dog” cure: hav­ing booze for break­fast to over­come the feel­ings of a hang­over from last night’s binge. It is the road to debt alco­holism, not the road to tee­to­tal­ism and recovery.

For­tu­nately, it’s a “cure” that is also highly unlikely to work, because the model of money cre­ation that Obama’s eco­nomic advis­ers have sold him was shown to be empir­i­cally false over three decades ago.

The first econ­o­mist to estab­lish this was the Amer­i­can Post Key­ne­sian econ­o­mist Basil Moore, but sim­i­lar results were found by two of the staunchest neo­clas­si­cal econ­o­mists, Nobel Prize win­ners Kyd­land and Prescott in a 1990 paper Real Facts and a Mon­e­tary Myth.

Look­ing at the tim­ing of eco­nomic vari­ables, they found that credit money was cre­ated about 4 peri­ods before gov­ern­ment money. How­ever, the “money mul­ti­plier” model argues that gov­ern­ment money is cre­ated first to bol­ster bank reserves, and then credit money is cre­ated after­wards by the process of banks lend­ing out their increased reserves.

Kyd­land and Prescott observed at the end of their paper that:

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 unfor­tu­nately they—and neo­clas­si­cal econ­o­mists in general—did bug­ger all about it. On the other hand, the Post Key­ne­sian group, of whom I am one, have con­tin­ued to try to con­struct mod­els of the econ­omy in which credit plays an essen­tial role.

I’ve recently devel­oped a gen­uinely mon­e­tary, credit-driven model of the econ­omy, and one of its first insights is that Obama has been sold a pup on the right way to stim­u­late the econ­omy: he would have got far more bang for his buck by giv­ing the stim­u­lus to the debtors rather than the cred­i­tors.
The fol­low­ing fig­ure shows three sim­u­la­tions of this model in which a change in the will­ing­ness of lenders to lend and bor­row­ers to bor­row causes a “credit crunch” in year 25. In year 26, the gov­ern­ment injects $100 bil­lion into the economy—which at that stage has out­put of about $1,000 bil­lion, so it’s a pretty huge injec­tion, in two dif­fer­ent ways: it injects $100 bil­lion into bank reserves, or it puts $100 bil­lion into the bank accounts of firms, who are the debtors in this model.

The model shows that you get far more “bang for your buck” by giv­ing the money to firms, rather than banks. Unem­ploy­ment falls in both case below the level that would have applied in the absence of the stim­u­lus, but the reduc­tion in unem­ploy­ment is far greater when the firms get the stim­u­lus, not the banks: unem­ploy­ment peaks at over 18 per­cent with­out the stim­u­lus, just over 13 per­cent with the stim­u­lus going to the banks, but under 11 per­cent with the stim­u­lus being given to the firms.

The time path of the reces­sion is also greatly altered. The reces­sion is shorter with the stim­u­lus, but there’s actu­ally a mini-boom in the mid­dle of it with the firm-directed stim­u­lus, ver­sus a sim­ply lower peak to unem­ploy­ment with the bank-directed stimulus.

Why does this model show that it’s bet­ter to give the money to the debtors than the lenders, in con­trast to the case that Obama was sold, that it’s bet­ter to give it to the bankers?

Because the “money mul­ti­plier” model is effec­tively a mechan­i­cal, sta­tic, equi­lib­rium model of the econ­omy. Give the banks excess reserves, and they will lend them to the pub­lic, which will hap­pily take on the debt. Once the reserves are fully lent out, the econ­omy is back to equi­lib­rium again.

In con­trast, my model is a dynamic, non-equilibrium one, where the “cir­cu­lar flow” of money and goods is prop­erly accounted for. In this sys­tem, you can think of the dif­fer­ent bank accounts in the sys­tem as like dams with pipes con­nect­ing them of vastly dif­fer­ent diameters.

When a credit crunch strikes, the pipes pump­ing the bank reserves to the firms shrink dra­mat­i­cally, while the pipe going in the oppo­site direc­tion expands, and all other pipes remain the same size.

If you then fill up the bank reserves reservoir—by the gov­ern­ment pump­ing the extra $100 bil­lion into it—that money will only trickle into the econ­omy slowly. If how­ever you put that money into the firms’ bank accounts, it would flow at an unchanged rate to the rest of the economy—the workers—while flow­ing more quickly to the banks as well, reduc­ing debt levels.

So giv­ing the stim­u­lus to the debtors is a more potent way of reduc­ing the impact of a credit crunch—the oppo­site of the advice given to Obama by his neo­clas­si­cal advisers.

This could also be one rea­son that the Aus­tralian expe­ri­ence has been bet­ter than the USA’s: the stim­u­lus in Aus­tralia has empha­sized fund­ing the pub­lic rather than the banks (and the model shows the same impact from giv­ing money to the work­ers as from giv­ing it to the firms—and for the same rea­son, that work­ers have to spend, so that the money injected into the econ­omy cir­cu­lates more rapidly.

This model can explain some aspects of the cur­rent US data that are inex­plic­a­ble from the con­ven­tional, neo­clas­si­cal point of view—the key para­dox being that while base money (“M0”) has been increased dra­mat­i­cally, there has been almost no move­ment in broader mea­sures of money (“M1” and “M2”). If the money mul­ti­plier argu­ment were cor­rect, the increases in M1 and M2 would have been mul­ti­ples of the increase in M0, as Obama was led to expect.

In fact, the expan­sion in M0 has been met by a fall in the credit-generated com­po­nent of the money sup­ply: since M2 includes all of M1 and M1 includes all of M0, this is clearer when we sub­stract the double-counting out. M1 has actu­ally con­tracted almost as much as M0 has expanded, while the expan­sion in M2 has been less than a third the size of the growth in M0.

The “money mul­ti­plier” has also collapsed—a mys­tery from a neo­clas­si­cal point of view, but entirely pre­dictable from the “endoge­nous money” perspective.

Obama has been sold a pup by neo­clas­si­cal eco­nom­ics: not only did neo­clas­si­cal the­ory help cause the cri­sis, by cham­pi­oning the growth of pri­vate debt and the asset bub­bles it financed; it also is under­min­ing efforts to reduce the sever­ity of the crisis.

This is unfor­tu­nately the good news: the bad news is that this model only con­sid­ers an econ­omy under­go­ing a “credit crunch”, and not also one suf­fer­ing from a seri­ous debt over­hang that only a direct reduc­tion in debt can tackle. That is our actual prob­lem, and while a stim­u­lus will work for a while, the drag from debt-deleveraging is still present. The econ­omy will there­fore lapse back into reces­sion soon after the stim­u­lus is removed.