But we're not talking about the early industrial age. We're talking about today.

I thought the article was actually lacking in that it didn't mention overproduction. This is the classic bane of industrialized economies: unsold inventory, rising unemployment, falling or negative inflation, falling asset yields, etc. The article describes China gaining the symptoms of overproduction.

And, as many writers both pro- and anti-capitalist have noted, it tends to require a solution extrinsic to the market (think of Paul Krugman's "babysitting co-op parable"). To make things profitable again, someone or something has to purchase the excess inventory at some market-clearing price. Nominal private debts also need to be cleared from the market, allowing general price levels to adjust.

Now here's the tough bit: overproduction is not the same as recession. Recession comes from malinvestment, and once capital is reallocated to genuinely productive uses it goes away. Some stimulus, especially infrastructure investment, also helps, but recessions are a natural part of the business cycle that end more or less on their own. Society just has to withstand the economic "bad weather".

Overproduction crises, however, also known as depressions, don't go away on their own. Apparent malinvestment in an overproduction crisis is only a symptom, not a cause.

If I could put forth a hypothesis, the actual cause is that one of the two big pillars of production has gotten way out of price-balance with the other. Those are labor and capital. In a healthy economy, labor is costly and productive, encouraging capital investment to get the most out of limited human labor-power. Capital, in turn, is well-priced for wide distribution and developed, increasing general productivity.

Knock out either of these, and you've got problems. If labor is too cheap compared to capital (ie: right now and the Great Depression), then you'll get overproduction: aggregate demand from labor wages won't exist to purchase the products of capital, and they rot. You also get low productivity growth, because it's that much cheaper to just employ humans at menial tasks (see: the service industry) rather than to automate.

(As an aside, computing is currently one of the few healthy fields: labor is expensive and capital (ie: FLOPs as well as investment dollars) is cheap, encouraging massive capital investments and productivity growth.)

If you knock out capital by making it too expensive or unavailable when labor is strong, you get stagflation. The capital-owning class and the working class bounce between themselves the price shock in the cost of investment or technological capital. So unemployment gets high (because where's the capital to employ people?) while price levels rise in general (because labor and consumers refuse to "take one for the team").

So to get rid of an overproduction crisis, what we actually need to do is increase the cost of labor, which feeds into aggregate demand to purchase down inventory and increase the marginal profitability of capital investments. I'd say the best way to do this is usually by either direct strengthening of workers (ie: allow labor to organize, shorten the work-week, other stuff like that), or by infrastructure investment (which employs laborers as well as yielding positive capital externalities).

Wasn't familiar with Krugman's "babysitting co-op parable". Interesting.

I'm not convinced that depression is strictly an issue of overproduction -- it's more a vapor-lock in which the currency pump between production and consumption seizes up. The Friedman approach is to encourage production, the Keynesian approach is to stimulate demand. Calling insufficient demand the same thing as overproduction doesn't quite seem right to me.

I've got to reflect on your cheap vs. expensive labor comments as well. The immediate problem I see is that expensive labor, in a world in which either capital or labor is free to migrate, results in either in-migration of cheap labor, or capital flight to less expensive markets. Which suggests possibly putting some controls on both, or addressing the labor costs part.

I suspect we're agreeing in at least part on the idea that you've got to have currency flows on both sides of the production/consumption cycle.

I'd also argue that right now both capital (in the form of interest rates) and labor (in the form of wages) are relatively inexpensive in the US. The dual problems are of stimulating demand, and on generating a real return on investment (equity or debt).

Overproduction is an issue insofar as (mostly) consumer goods are how we manage to generate fiscal flows. People buy goods for money. However if they either don't buy goods, or prices are too low to generate sufficient flows, you end up in a cycle where you're cranking out crap goods which people don't particularly need to consume but are strongly encouraged (by advertising, moral suasion, societal pressure) to do so: the "consumer economy".

We worked our way out of this the last time (Great Depression) with WWII.

Definitely. I can say that at least right now, one of the major problems is that labor is "penned in" by national borders, but capital isn't. The labor side of things has gotten too weakened because capital can arbitrage itself to seek optimum pricing, and labor can't.

So yes, to a certain degree I'm in favor of capital and/or trade controls simply because I think a real one-world open-borders policy cannot possibly work right now, and free trade/capital flow without free movement of labor is just an arbitrage opportunity waiting for eager capitalists.

I'm not convinced that depression is strictly an issue of overproduction -- it's more a vapor-lock in which the currency pump between production and consumption seizes up.

True enough. A "real depression" also seems to have another major component: a weight of private debt that the nonfinancial economy cannot reasonably carry. This prevents price and currency adjustments from correcting the other imbalances naturally.