Notes on Overinvestment
March 9, 2000
Department of Economics
Loyola Marymount University
In personal correspondence, Professor Brad DeLong of the University of California, Berkeley, writes: "I have never yet seen an 'overinvestment' cycle: wrong kinds of capital, yes; deficient demand, yes; inflation control, yes; but not 'overinvestment.'" This is a good point, common to Keynesian economics. It requires a reply.
Overinvestment is defined as too much investment from the perspective of capital as a whole (the capitalist class): there's either too much to maintain a sufficiently high profit rate or too much to allow a stable growth process and thus the stability of profit income. This "or" corresponds to the two sorts of overinvestment that I have found for the rich countries (in my research):
(I) A kind of overinvestment relative to supply occurs when fast accumulation (associated with high demand for labor-power) pulls up wages, which squeezes profits. This is associated with raw material prices being pulled up and excessively high capital/output ratios. It happens in what I call a "labor scarce economy" such as the US economy in the 1960s and helps explain the fall of the profit rate in the late 1960s. Of course, the story is more complicated: expansionary fiscal policy (due to Lyndon Johnson's war against Vietnam, his fear of domestic opposition, and the Federal Reserve's going along) kept the boom going for a long time, so that profit-rate-depressing imbalances could accumulate.
Bob Brenner's recent book points to another dimension of this: increasing international competition among the rich capitalist countries meant that there was intensified competition over profits which encouraged international overinvestment. This meant that for all of the major countries to use their capital goods at the full-capacity level would have meant even more of a wage- and raw-material-price-squeeze on profits than actually happened. It also puts a lid on each country's export price level, limiting the ability to restore profitability by raising prices.
On this kind of overinvestment, see my article in the Eastern Economic Journal, July-August 1997.
(II) The second main kind of overinvestment is relative to consumer demand. In a "labor abundant economy" such as the US in the 1920s or the 1990s/2000s, wages lag behind productivity so that profits and profit rates soar. As a result, working-class consumption from current income falls behind total production. In order to allow the realization of rising profit rates, the gap between production and working-class consumption from income can be filled by (a) rising investment as a percentage of GDP; (b) rising government deficits; (c) rising trade surpluses or falling deficits; (d) rising luxury spending by the rich and their neighbors on the income distribution curve; and/or (e) rising working-class consumption based on debt accumulation.
In the 1920s, (a), (d), and to a smaller extent, (e) dominated, so that the profit rate could continue to rise; both (c) and (b) were ruled out. In the 1990s/2000s, the story is very similar, especially since (c) and (b) are both happening in reverse, depressing the economy. The big difference, it seems, is that working-class debt accumulation is larger. However, this goes along with a cyclical investment surge.
Rising investment and debt can allow the realization of a rising rate of profit for quite awhile (in theory), but they correspond to an accumulation of imbalances: increasing productive capacity, implying the need for even more investment to realize a high profit rate, plus rising debt and debt-service payments. Further, investment spending is notoriously more flaky (i.e., volatile) than consumption spending, since it is easy to delay and is based on guesses about future profitability. (If all other firms are investing, that encourages each to do it; similarly, if everyone stops investing, the competitive drive to invest fades.) Further, increasing consumer debts implies that interest payments rise relative to income, encouraging bankruptcy.
(On the above, see my 1994 article in RESEARCH IN POLITICAL ECONOMY. See http://clawww.lmu.edu/~JDevine/depr/D0.html or http://clawww.lmu.edu/~JDevine/depr/nushortdepr.html.)
(III) There's a third kind of overinvestment for poorer countries, such as those of East Asia, pursuing an export-oriented growth path. Competing with each other for the same rich-country markets, they overinvest in the sense that if their capacity were fully used, it would mean even lower prices for them (lower exchange rates). Because of the inelastic demand for their products (i.e., for the products of their economies all added together), falling prices or exchange rates don't help them. This is similar to the phenomenon of overinvestment by small farmers (petty capitalists), who regularly invest too much (in the face of inelastic demand) and drive many of their members into bankruptcy. Falling prices (exchange rates) simply make debt service more expensive. Since the investment boom usually corresponds to debt accumulation, this is crucial.
DeLong continues: " after all, enough monetary expansion can push interest rates low enough to make any investment that is at all productive profitable..."
Ricardo Caballero's research (presented at the ASSA convention in January 2000) points to the real phenomenon of the vertical IS curve due to excessively idle capacity (along with excessive corporate debt and pessimistic expectations). This is what I call the Depression IS curve: falling interest rates do not induce much if any investment under these conditions, so that there's little increase in GDP. This situation can be reinforced by excessive consumer debt, consumer pessimism, and overbuilding in the housing market. (This is a situation that Leijonhuvfud would call "outside the corridor.") Real interest rates cannot fall below zero unless expected inflation rises significantly (as Paul Krugman points out). So, absent severe inflation, even a steep (i.e., non-vertical) IS curve may not intersect the LM at full employment. (Note that even before zero, there's a floor on interest rates because of the liquidity value of money as an asset.) This implies that monetary policy is impotent.
In this kind of situation, expansionary fiscal policy is needed. But of course President Coolidge -- oops, I mean Clinton -- wants to commit the US to "paying down" the government debt by 2013. Most of the Republicans will likely go along, of course, though they might push for some gifts to the rich (just as the Democrats will push for mild gifts for the middle class). So it will take a big crisis to shake the government to change course and expand the economy. Maybe a war would change policy …