Steve Keen’s DebtWatch No 23 June 2008

RBA Assistant Governor Guy Debelle and I spoke at a conference on Subprimes in Adelaide last month. One aspect of my analysis that Guy queried was my emphasis upon the Debt to GDP ratio. He noted that this appeared suspect, because it was comparing a stock (the outstanding level of debt) to a flow (annual GDP).

It’s a valid point to make. The engineer-turned-economist Mickal Kalecki once caustically observed that “economics is the science of confusing stocks with flows”, and I’m a stickler myself for not making that mistake. So making a song and dance about a stock to flow comparison like debt to GDP has to be justified by a sound argument.

The most accessible argument is an analogy to global warming. Just as the growing level of C02 in the atmosphere is evidence that the ecosystem is not coping with the (relatively tiny) additional volume of carbon dioxide human activity is adding to the biosphere, the accumulation of debt relative to income is evidence that the economy is not coping with the (relatively large) volume of debt being generated by the financial system to finance speculative purchases of assets.

There are at least five strong similarities between this ecological issue and the economic one of accumulating debt:

  1. The environment has an established “carbon cycle”, by which the emission of carbon dioxide into the atmosphere by animals, etc., is balanced by its re-absorbtion by plants, etc. This process is augmented by many other factors–such as the explusion of CO2 by volcanoes–and, on a geological time frame, far from stable; but over the period of human existence, the atmospheric concentration was below 300 parts per million (ppm) until industrialisation began.
  2. The fact that the concentration has risen from under 300 ppm to over 380 ppm in the past century shows that the planet’s natural carbon-processing cycle is not coping with the extra CO2 added by human industrial and agricultural activity. Global Warming is the most apparent manifestation of this increased stock of CO2.
  3. Even if all human industrial and agricultural activity stopped today, it would take decades for the additional CO2 we have added to the atmosphere to decline to pre-industrial levels. At present, the existing natural carbon cycle is only means to reduce that accumulation, and it could take longer to reduce the build up than it took to accumulate it, since we have damaged many of the natural “carbon sinks” in the ecosystem.
  4. Persisting with present or higher levels of CO2 will cause the ecosystem to undergo both profound and uncertain change, some aspects of which can be inferred from past geological data.
  5. There are feedback effects that mean any climate-altering effects of raised CO2 levels may be further amplified. For instance, the rise in temperature has reduced the area of ice in the North Pole, leading to reduced reflection of sunlight (since ice reflects most light while water mostly absorbs it), and further increasing global temperatures.

Similar propositions can be put with respect to debt:

  1. The economy has a natural capacity to process debt. Borrowing by businesses to finance investment can lead to new products whose sale enables the businesses to make a profit and repay debt over time (Schumpeter’s Theory of Economic Development gives perhaps the best explanation of this “natural debt cycle”);
  2. The fact that debt levels are rising with respect to income is a sign that the economic system is not coping with the level of debt being generated today;
  3. Even if all borrowing stopped today, it would take decades for the additional debt we have accumulated to be reduced to pre-debt bubble levels. The only way that debt levels can be reduced is if income is redirected from either consumption or investment into debt reduction. This is the key reason that the debt to GDP ratio is important: it tells us how much of income would be needed to reduce debt, and how long such a reduction would take. However the process of reducing debt will itself reduce income to some degree, since money that would otherwise have gone into investment will now simply be used to pay down debt levels–and incomes and employment will fall as a result;
  4. Persisting with the level of debt we have now will mean a profoundly different and uncertain economic environment. The proportion of income needed to service debt will remain at levels that have only ever been experienced in the past during Depressions.
  5. Finally, there are feedback effects in the economy that can mean debt to GDP ratios fall even when, ultimately, borrowers try to reduce their exposure to debt. The worst such effects are: direct reductions in investment as retained earnings are used to pay down debt rather than invest; indirect falls in investment as falls in consumption reduce cash flow and depress both earnings and expectations; and falling prices if deflation sets in, as it did during the Great Depression when prices fell as much as 10% per year.

There is also an academic economic argument that can be made about the importance of the debt to GDP ratio, deriving from Minsky’s “Financial Instability Hypothesis”. In this theory, a rising ratio is both a prediction of the model, and a force leading to greater financial instability and possible economic breakdown in a Depression. But I’ll leave a full discussion of this for a future Debtwatch.

Observations on the Data

The most recent data imply that the turnaround in debt to GDP may finally be starting. The debt to GDP ratio fell last month, from 165.25% of GDP to 165.2%. It’s not a lot, and it may still return to its 44-year long upward trend; but it is the first time in fifteen years that the ratio has fallen.

The contributor was an “unexpected” fall in the rate of growth of business borrowing (these things always seem to be “unexpected”). Though it still increased, it grew at a slower pace than GDP–as did personal borrowing. The grwoth in mortgage debt, on the other hand, continued to outstrip GDP (see Tables 1 and 2 below for details).

While in one sense this slowing down in debt growth is a good thing–in that an unsustainable trend may finally be coming to an end–it also may presage very tough economic times ahead. One aspect of my focus on the debt to GDP ratio is the contribution that change in debt then makes to aggregate spending. Aggregate demand in the economy, for everything from commodities to net asset transfers, is the sum of both income (GDP) plus the change in debt.

In a well functioning economy, that shouldn’t be much relative to GDP itself–and it wasn’t in the 1950s and 1960s. Then, the annual change in debt contributed no more than 4% of total demand.

But as the debt level rises, the change in debt can become an extremely large and volatile component of aggregate spending. That clearly is what happened from 1970 onwards: the annual change in debt began to contribute substantially more than 4% of total demand.

And it was highly volatile: notice the slump in its contribution from over 10% to just 4% in 1973-75 that coincided with the collapse of the Whitlam Government, and the fall from a peak of almost 14% of aggregate demand to minus 1.5% during “the recession we had to have”.

Today, the annual change in debt is the source of over 19% of aggregate demand. Should that turn around–as it must even to stabilise the debt to ratio at its current historically unprecedented level–then demand in the economy could “unexpectedly” evaporate.

The full Debtwatch Report, with all the graphics, is available in PDF format here.