Why We Should Not Trust The Fed's Capacity Utilization Numbers

by: Harini Dedhia


There are three separate government agencies collecting different components.

Unequal survey term for different components.

Varied sample and sample size for the production and capacity numbers.

Capacity utilization is not as low as the Fed claims it is.

The investment community watches with great avidity when the capacity utilization number is released by the Federal Reserve Board every month hoping it serves as an indicator of the nation's economic health. However, the numbers do not quite hold up to this undeserved adulation and here's why.

The monthly aggregate capacity numbers are computed by interpolating an annual capacity index that is computed using the FRB's production index and the Census Bureau's Survey of Plant Capacity (SPC) (which is monitored jointly by the FRB and the Department of Defense). Two surveys with varying samples and different authorities sponsoring it are hence producing an aggregate annual number that is then interpolated to give us monthly data. If this itself was not problematic enough, here are the reasons why the Census Bureau's annual survey of plant capacity cannot be trusted:

  • The annual survey measures the fourth quarter rates of capacity. The survey respondents are asked to assume a product mix that is typical of their production during fourth quarter. The monthly iterations produced for the Fed's final index using their industrial production numbers (IP) divided by this SPC number therefore are based on numbers coming out annually and are dependent on fourth quarter utilization rates. 'Abnormal changes' during the following months would be missed even in the Fisher index adjustments to produce the monthly numbers.
  • The SPC, jointly sponsored by the FRB and the DOD measures utilization of around 17,000 plants. There are approximately 330,000 manufacturing establishments in the nation. Thus this survey only surveys about 5% of the total establishments.
  • The survey respondents are asked to assume that labor, utilities and materials are fully available. Prior to 1996 survey, the respondents were asked to assume all these supply variables to be equal to the maximum usage level attained in five years. Thus, for example, labor shortage in the Permian resulting in lower capacity usage would demonstrate economic slack and give a negative price pressure signal when the reality is quite contrary. The number of shifts and hours of plant operation are assumed to be sustainable under a 'normal' and a 'realistic' work schedule.
  • The sample is revised once every five years with a new probability sample drawn from the census of manufacturers. However other checks on telephone and mail respondents are very limited. The survey, despite being termed as "mandatory," has a response rate of 80%. The only annual change in the sample is undertaken to reflect new plant births.
  • Each industry is treated independently and establishments are selected with a probability proportionate to size. The aggregated utilization rates are then derived from capacity weighted aggregates of industry utilization rates. However, certain industries that are identified by the DOD as "priority industries" are sampled more heavily than the others. This leaves an inherent bias in the dataset.
  • Plants that are idle but not officially permanently closed are also included in the capacity under the SPC.
  • Services such as publishing and printing are considered as manufacturing and thus incorporated in this survey. However, industries such as logging are excluded from the survey as it falls under the purview of the Department of Agriculture. This is misleading.

The redundancy of such a metric is hence very evident. While it can be used as a good lagging indicator of trend, it cannot accurately quantify the slack in the economy. The defensive play on equities based on this non-existent slack thus leads to a low growth play for many. I believe that the US is quickly running out of economic slack and hardly think that the Fed's utilization number begins to capture it for the reasons stated above. Being bearish or defensive with US equities based on this pop-notion of 'slack' would certainly leave investors out of a growth cycle period.

Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it. The author has no business relationship with any company whose stock is mentioned in this article.