It’s just a flesh wound…”

Flattr this!

It seems we’ve moved from Stan­ley Kubrick to John Cleese. Rory Robertson’s reply to my “Rory Robert­son Designs a Car” post reminds me of one of my many favourite scenes from Monty Python, the fight between King Arthur and the Black Knight:

King Arthur: [after Arthur’s cut off both of the Black Knight’s arms] Look, you stu­pid Bas­tard. You’ve got no arms left. 

Black Knight: Yes I have. 

King Arthur: *Look*! 

Black Knight: It’s just a flesh wound…

Most of Rory’s com­men­tary in his newslet­ter has been repro­duced by Christo­pher Joye in an amus­ing “ring­side report” (Keen vs. Robert­son: Round V) on the Busi­ness Spec­ta­tor (inci­den­tally, Chris’s post includes an excel­lent dig at the RBA’s per­for­mance in recent years; this is well worth a read in its own right).

Tak­ing Chris’s extract as a guide, it seems that Rory’s entire con­sid­er­a­tion of my post boils down to this:

**Need­less to say, Dr Keen does not accept the assess­ment that a “school­boy error” lies at the heart of his pes­simistic fore­cast of a 40 per cent drop in Aus­tralian home prices. But instead of address­ing the key point that debt ser­vic­ing just got much eas­ier for most home-buy­ers, Dr Keen responded by invent­ing a silly story about cars and fuel con­sump­tion — to make a point that com­pletely missed the point…

Does not accept the assess­ment”? Do we have a Dead Par­rot talk­ing here, as well as an arm­less Knight? The point of my post was that Rory’s argu­ment that com­par­ing Debt to GDP is a “school­boy error” (“like com­par­ing apples with oranges”) was itself a school­boy error that betrayed the depress­ing lack of under­stand­ing that most neo­clas­si­cal econ­o­mists have of dynam­ics. In engi­neer­ing and many other prop­erly dynamic dis­ci­plines, stock to flow comparisons–like com­par­ing Debt to GDP–abound. Far from being a “school­boy error” to make them, it’s a “haven’t been prop­erly edu­cated at uni­ver­sity” error to deride them.

They can be done in error, sure–when the result­ing mea­sure has non­sense dimen­sions, or is irrel­e­vant to the issue at hand. Com­par­ing Debt to GDP isn’t an instance of either error, since as I showed in that post, the result­ing dimen­sion is “Years”. The ratio mat­ters because it tells you how long it would take to reduce debt to a given tar­get, if a given per­cent­age of income was devoted to repay­ing it.

The cur­rent answer is 1.59 Years, if all GDP was devoted to debt repay­ment (which can’t hap­pen of course–5% of GDP p.a. is a more likely delever­ag­ing rate) and if the tar­get was a zero % debt ratio (which it wouldn’t be–the 1950–70 range of 25–50% is more likely), and if reduc­ing debt didn’t affect GDP (which unfor­tu­nately ain’t the case–there will be many dam­ag­ing pos­i­tive feed­backs from reduc­tions in debt to reduc­tions in GDP).

And for Pete’s sake, a “stock to flow” com­par­i­son was the linch­pin of Friedman’s Mon­e­tarism, as a blog mem­ber here pointed out:

cheap­bas­tud said, on March 16th, 2009 at 1:22 am:

uhhhh, isn’t VELOCITY from the equa­tion of exchange a stock/flow ratio (in this case it’s a flow/stock ratio)?


Maybe I’m mak­ing a hor­ri­ble school­boy error or maybe Mr. Robert­son doesn’t know wtf he’s talk­ing about.

Spot on. The “veloc­ity of money” is the num­ber you get from divid­ing nom­i­nal GDP ($/year) by the money stock ($). Its dimen­sion is 1/years, so that the inverse of the ratio tells you how many times the money stock turns over in a year. The for­lorn attempt to prove this was a con­stant was Friedman’s key research objec­tive, since if V wasn’t a con­stant then much of the Quan­tity The­ory of Money was inval­i­dated.

Now I doubt that Rory is going to accuse Fried­man of com­mit­ting a “school­boy error” here (though in truth Fried­man is guilty of so many that there should be a Fried­man Prize in School­boy Errors–and  vir­tu­ally every year it could be awarded jointly with the Nobel Prize in Eco­nom­ics). So why accuse me?

In my long expe­ri­ence with attempt­ing to debate eco­nom­ics with neo­clas­si­cal econ­o­mists, I have become accus­tomed to an often irrel­e­vant and fre­quently false point being raised, after which dis­cus­sion is ter­mi­nated. The point of rais­ing the point is not to engage in debate, but to shut it off. So too with this patently false argu­ment that, because I use a “stock to flow” ratio, the remain­der of my argu­ments can be ignored.

In itself, there’s noth­ing wrong with argu­ing this way–in a reli­gious debate. If you have two per­spec­tives, one of which sees a God as cru­cial to under­stand­ing the uni­verse, and another which doesn’t, then they’re always going to argue past each other. But eco­nom­ics isn’t sup­posed to be a religion–it had, at least until this cri­sis hit, the pre­ten­sion of being a sci­ence.

I has­ten to add that I don’t see this as delib­er­ate eva­sion by Rory, nor even nec­es­sar­ily con­scious evasion–and ditto for the many neo­clas­si­cal cor­re­spon­dents and ref­er­ees I’ve dealt with over the years. They can, and do, cope with debates within the con­fines of their own belief sys­tems; so if I was argu­ing that the NAIRU (don’t bother ask­ing what it is if you don’t already know–it’s not worth the effort of dis­cus­sion!) was 4% rather than 6%, or maybe even that prices were sticky down­wards rather than per­fectly flex­i­ble, I might get an argu­ment.

But when you effec­tively chal­lenge core beliefs–by argu­ing, for exam­ple, that equat­ing mar­ginal cost to mar­ginal rev­enue doesn’t max­imise prof­its (again, don’t bother, but if you must, check here)–you get a non­sense reply to shut the debate down.

In a true sci­ence, a sub­stan­tive point is either con­tested or con­ceded. Rory did neither–though to cut him some slack here, he might not have realised why I wrote my piece either. I didn’t give him any fore­warn­ing, so he was free to make a mis­taken inter­pre­ta­tion of why I wrote some­thing about him. Instead, whether he meant to or not, he ignored my main point, and changed the topic back to the house price issue .

From his point of view, I changed the topic, which in his post was house prices–his “stocks and flows” state­ment was just an aside. But in fact, if house prices had been all Rory had talked about in the newslet­ter to which I responded, I wouldn’t have both­ered writ­ing any­thing.

It was the “com­par­ing stocks to flows is a school­boy error” non­sense that inspired me to write some­thing (and I was also respond­ing to a reader’s request that I assist him in con­test­ing that spe­cific propo­si­tion). But Rory’s take is that I made my com­ment to dis­tract atten­tion from his argu­ment about house prices:

**Regard­less, there remains a large hole in Dr Keen’s analy­sis (big enough to fit a bus?). Barely six months ago, he was high­light­ing the uptrend in the house­hold sector’s inter­est-to-income ratio as the key force that would bring house prices crash­ing down.”

Quot­ing Keen: In 1990, ser­vic­ing mort­gages cost three cents of the house­hold dol­lar — now its 15 cents, even with lower inter­est rates. …This is because of the sheer size of the debt — that’s the pres­sure that’s going to be push­ing house prices down and it’s actu­ally the same kind of pres­sure that is in the US (see here; (The Age back in Octo­ber also reported: “…Mr Keen said the lower end of the [hous­ing] mar­ket was already col­laps­ing”. Really?)”

**Now that his debt-ser­vic­ing ratio has crashed towards 10 per cent from 15 per cent, Dr Keen has noth­ing to say on the mat­ter. Fur­ther­more, with the ratio now trend­ing lower, Dr Keen has stopped pub­lish­ing the debt-ser­vice chart that once was at the cen­tre of his analy­sis. (Between Novem­ber 2006 and May 2008 the debt-ser­vice chart was reg­u­larly pub­lished in Debt­Watch; for exam­ple, see Fig­ure 21 in the Feb­ru­ary, April and May 2008 reports, and Fig­ure 12 in the Novem­ber 2006 report, see here).

**Some­one unkind might won­der if Dr Keen is steer­ing clear of key facts that directly con­tra­dict the scary story he likes to tell. Pauline Han­son might be inclined to issue a “Please Explain”? In any case, con­trary to Dr Keen’s ill-informed claim in the quote above, the sit­u­a­tions in Aus­tralian and US hous­ing and mort­gage mar­kets are very dif­fer­ent, like chalk and cheese (see charts 4–15 in the attached PDF file).”

Well, actu­ally, no Rory. Firstly, I men­tioned two forces: the inter­est rate bur­den, and de-lever­ag­ing. True, the for­mer has fallen some­what; the lat­ter is as potent as ever, and only just begin­ning to click in here, while it’s dri­ving the col­lapse in the USA and Europe. I pub­lished two charts on that in the Dr StrangeLove post (they’re repro­duced at the bot­tom of this post too), but I wasn’t ignor­ing the inter­est rate issue either.

The inter­est pay­ment bur­den, while it has dropped sub­stan­tially cour­tesy of the RBA’s belated and pan­icked cuts to rates, is still at higher lev­els than at any time out­side the period 1989–1991–when rates were three times what they are now.

The rea­son I haven’t been pub­lish­ing these charts in Debt­watch is not that they no longer make the case I want, but because the reports were grow­ing too long and–given the soft­ware I was using to pro­duce them–the lay­out was becom­ing too messy. I’m work­ing with a few blog mem­bers to pro­duce a web-acces­si­ble inter­face to all the data that will get around this prob­lem ulti­mately, but it takes time to do this.

In the mean­time, here are some of those charts. Firstly, inter­est rates and the inter­est rate pay­ment bur­den as a per­cent­age of GDP: rates and the bur­den have fallen, and sharply, but still only taken us back to lev­els that applied when aver­age rates were 16% or higher between mid-1988 and early 1991. That’s hardly heaven.

How much fur­ther rates have to fall to return the inter­est rate bur­den to any­thing close to the aver­age since 1960 is indi­cated by this next chart. We’re still way above the aver­age bur­den for the last half cen­tury.

The aver­age inter­est rate used above is a weighted aver­age of busi­ness, mort­gage and per­sonal rates (and it prob­a­bly under­states the bur­den on busi­ness slightly, since I had to guess the lat­est figure–the RBA only updates busi­ness rates on a quar­terly basis, and I extrap­o­lated from the Sep­tem­ber fig­ure using the most recent gap between the 3 year fixed rate for small busi­ness and the vari­able rate for large busi­ness; this gap was the small­est it’s been in years, so the busi­ness rate is prob­a­bly higher than I guessti­mated here). Break­ing this down, and com­par­ing the busi­ness pay­ment bur­den as a per­cent­age of Gross Oper­at­ing Sur­plus and the house­hold rates as per­cent­ages of House­hold Dis­pos­able Income yields the fol­low­ing chart:

Thus while the bur­den on busi­ness is sub­stan­tially below what it was in 1990 (when the RBA’s rate hike to attempt to tame the asset bub­ble back then had a crip­pling impact on busi­ness) the bur­den on house­holds now is still more than 4% higher (as a pro­por­tion of dis­pos­able income) than it was in 1990.

The rea­son for this, of course, is the dra­matic increase in mort­gage debt over the last twenty years. Ana­lysts who believe that house prices will always rise focus just on that datum itself. I’ve argued from a Hyman Min­sky, “Finan­cial Insta­bil­ity Hypoth­e­sis” point of view, that this trend of ris­ing house prices only occurs because the debt bor­rowed to buy houses has risen faster still.  The next chart, which indexes both mort­gage debt and house­hold prices to 100 in 1996, illus­trates that point:

Finally, there are of course two forces that deter­mine the inter­est repay­ment burden–the rate of inter­est and the level of debt (three actu­ally if one looks at the real bur­den, but I couldn’t find that chart in a hurry so I’ll  leave it for another day). If you plot the level of debt as a pro­por­tion of GDP on a hor­i­zon­tal axis, and the inter­est rate on the ver­ti­cal, then you can show com­bi­na­tions of Debt to GDP ratios and inter­est rates that have an equiv­a­lent out­come in terms of the inter­est rate bur­den: thus a com­bi­na­tion of a Debt to GDP ratio of 40% and a nom­i­nal inter­est rate of 20% has the same impact as a bur­den of 400% and an aver­age rate of 2%.

Check­ing the actual time path of the inter­est rate bur­den on this chart, I sur­mised that a spec­u­la­tive boom seems to occur when­ever the bur­den falls to about the 8% level, whereas the max­i­mum bur­den we’d ever expe­ri­enced was 16.7% in 1990, with a debt ratio of 83% and an aver­age inter­est rate of 19.7%.

Also, inter­po­lat­ing from the mean gap between the cash rate and aver­age inter­est rates of 3.3%, it appeared that the debt ratio at which the min­i­mum debt bur­den would be that “good times” level of 8%, was 240%: if the debt to GDP ratio ever hit that level, then there was no way the “good times” could ever come back. That analy­sis is shown in the next chart. Though we’ve retreated from the “max­i­mum pain” line of 16.7%, we’re still well above the “good times” level of 8%–it would in fact take a fur­ther 3% fall in inter­est rates to take us back there, if there was no change in the debt ratio.

So there isn’t much room for rate cuts to reflate the economy–a 3% fall in aver­age rates would require the cash rate to fall to 0.25%, and all of the rate cut to be passed on. That head­room would fall even fur­ther if that “school­boy error” Debt to GDP ratio rose any fur­ther.

Thus the inter­est rate cuts have reduced the pain of debt ser­vic­ing, but they haven’t reduced them to any­thing near the com­fort­able lev­els of the pre-1980s. And the main prob­lem of debt-delever­ag­ing remains, and will be the main fac­tor dri­ving the econ­omy down, as the con­tri­bu­tion that change in debt makes to aggre­gate demand plunges. That effect is already patently obvi­ous in the US data:

And the first signs of the same process are now turn­ing up in the Aus­tralian data, with the most recent “unex­pected” increase in the unem­ploy­ment rate:

The impact of de-lever­ag­ing is the main force that I see dri­ving us into Depres­sion, and tak­ing house prices down in the process. There may well be a fil­lip to the bot­tom end of the hous­ing mar­ket out of the Government’s ludi­crous boost to the First Home Buyer’s Grant, but the weight of delever­ag­ing will, I expect, soon tell against that.

And Rory, let’s lighten up here please. My Dr StrangeLove post was meant to make in a comic fash­ion a point that obvi­ously hadn’t got­ten through via seri­ous dis­cus­sion: that stock to flow com­par­isons are, if done cor­rectly, legit­i­mate aspects of analysing a dynamic sys­tem. Con­cede that point, and I’ll tickle you with my next sword thrust, rather than slic­ing your legs off.

Over to you, Mr Joye, for the ring­side com­men­tary.

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.
Bookmark the permalink.
  • Hi Shadow,

    I’m aware of Battelino’s com­ments there–and I wrote a post on it when it was pub­lished.

    I’ve also recently com­mented on whether GDP is com­pa­ra­ble to credit in two other posts–one a satire on Rory Robert­son, the other a fol­low up to Rory’s reply (I see you’re replied to the lat­ter).

  • Jim

    Hi Shadow,
    Since I am not an econ­o­mist I might be able to explain it more sim­ple terms. Like you, I was con­fused about the same thing up until Steve pub­lished the Inter­est Rate-debt Trade­off chart above. Like me, you see the real issue as the “repay­ment bur­den” (inter­est on the debt). The chart shows that as the ratio of debt to GDP increases the level of inter­est rates must fall to main­tain the same repay­ment bur­den. This can not keep going on for­ever. At some point inter­est rates will hit zero and then the repay­ment bur­den will start mov­ing up as the debt to GDP moves up. You might argue that we have not reached that point now. Well, I also made the same mis­take as you again. The penny dropped for me when I realised you can not look at your own case in iso­la­tion. If the econ­omy is pro­duc­ing less then less peo­ple will be employed. This means that should you decide to sell your prop­erty it will be harder to sell, so you will have to reduce the price and you end up los­ing money. You might say to me that you don’t plan to sell your prop­erty, but if you get unem­ployed then you might be forced to sell it.

    At the end of the day, you need to look at every­body with loans in a col­lec­tive way rather than look­ing at one case in iso­la­tion. I hope this clears things up for you.