Capacity utilization has tended to be a good coincident indicator of the economy. However, while previous peaks have led us to at or above 85% capacity utilization, the 2007 peak gave us a capacity utilization in the lower 80% range. With current capacity utilization at 78%, there are two possible interpretations: (a) we are either in the early to middle innings of an economic recovery. Or (b) if we double dip, we will have peaked at a level of capacity utilization that’s lower than the 1970s–1980s head fakes (Figure 1).
But how much does capacity utilization matter? If we take a look at our level of private nonresidential investment, with our current cyclical low at levels comparable to the 1950s (Figure 2), at least you can see that we’re still investing in the future.
However, the denominator, GDP, is currently at a depressed level. Second, another issue is obfuscated when you only look at the higher level data: deindustrialization.
There’s been a lot of talk lately about the deindustrialization of America. As we have shifted to a service economy, we have also shifted the nature of our capital expenditures. Instead of investing in factories, we have been increasingly investing in equipment and software (Figure 3).
But the NIPA tables don’t tell us whether the equipment we invest in is a computer or factory equipment. One possible way we could impute the division is to look at the life of the assets purchased.
If we look at the consumption of capital versus the investment of capital (Figure 4), there has been a secular increase in the consumption of capital versus investment in capital. This could reflect the increased share in investments such as computers, which economically devalue according to Moore’s Law whilst older equipment’s economic depreciation is over a longer time horizon.
A more concrete way to think about this is when you jettison the manufacturing part of the business but retain the higher value activities in the United States, (think of Apple’s “Designed in California, Made in China” strategy), then you have workers sitting in front of computers instead of machines. Those computers halve in value over two years whereas factory equipment would have devalued based on wear and tear over horizons of 5, 10, even 30 years.
So, as American companies have been able to shift to capital light models, does capacity utilization still matter? If it does, then when you realize that the lower capacity utilization is already happening at a lower investment level, then the economy is in worse shape than you would think.
But if you think that we have a viable path to recovery that can be based on a knowledge and service economy, then maybe the vicissitudes of capacity utilization needn’t keep us up at night.