In this Section:

·   Survey Design

·   Data Collection

·   Compilation of Data

·   History of Survey Program

Survey Design

Target Population:

The target population for the Quarterly Survey of Plant Capacity Utilization (QPC) is defined as all domestic establishments with five or more employees that are classified in any of the 2017 North American Industry Classification System (NAICS) industries that make up the 93 industry groups specified by the Federal Reserve Board (FRB), one of the QPC survey sponsors. The 93 industry groups defined by the FRB are the levels of publication and are comprised predominantly of 4-digit NAICS industries or combinations of industries in the manufacturing sector and the publishing subsector of the information sector.

Sampling Frame:

The sampling frame for the current QPC sample, introduced in the first quarter of 2020, was constructed using information extracted from the Census Bureau’s 2018 Business Register, with updated information from the 2017 Economic Census. This sample frame contained approximately 189,000 manufacturing establishments and 8,200 publishing establishments located in the United States having five or more employees and classified in any of the 93 industry groups described above. The 2020 QPC sample was selected on a 2017 NAICS basis, which reflects any changes to industry definitions resulting from the 2017 NAICS revision.

Sampling Unit:

The sampling unit for the QPC is the establishment.  Sample estimates are produced for the QPC from a sample of roughly 7,500 establishments.

Sample Design:

The sample for the QPC is selected using a probability-proportional-to-size (pps) sample design based on the assigned measure of size. Sampling probabilities for the quarterly survey are assigned proportionate to a measure of size that approximates annual total value of shipments. There are some establishments specified by the QPC survey analysts to be included in the sample as predetermined certainties. These establishments are flagged up front and included in the sample with weights of 1. We also impose a minimum probability rule, which results in a maximum sample weight (inverse of the minimum probability). This minimum probability rule reduces the risk of exceedingly large sample weights that can adversely impact sample estimates and variances. For the QPC, the minimum probability is 0.02, resulting in a maximum sample weight of 50.

Sample allocation is determined by the priority industry requirements specified by the FRB across the 93 industry groups designated as publication levels. Each of these 93 industry groups is designated as high, medium, or low priority and is comprised of predominantly 4-digit NAICS industries or combinations of NAICS industries. Sampling is controlled under tighter CV constraints for the high priority industry groups. Each of the 93 industry groups is sampled independently to satisfy the total sample size constraint. The sampling methodology ensures that the allocated sample size for each industry group is exactly realized and yields the desired total sample of 7,500 establishments.

Frequency of Sample Redesign:

Every 5 years (implemented two years after each Economic Census, which is conducted for years ending in ‘2’ and ‘7’).

Sample Maintenance:

Since the QPC sample is redesigned every five years, there is a need for sample maintenance in the intervening years. Each survey cycle, establishments are lost through sample attrition, so something needs to be done in order to maintain the desired total sample size of 7,500 establishments. Therefore, in each intervening year, a birth sample is selected in order to accurately reflect the universe for a given survey year and to offset the effects of this sample attrition each survey year. The target number of establishments selected in each annual birth sample is determined by the attrition rate from the previous survey cycle. Similar to the full sample selected every five years, each birth sample is allocated across the 93 industry groups published in the QPC based on the respective establishment attrition rates in each industry group. This ensures that respective industry group sample sizes are maintained, while also maintaining the total sample size of 7,500 establishments.

Data Collection

Data Items Requested and Reference Period Covered:

Operational status for current quarter.

Actual, full, and emergency production data are collected for the current and prior quarters, while actual and full production comparison checkbox information and work pattern data are collected for the current quarter only.

The survey questionnaire can be found at https://www.census.gov/programs-surveys/qpc/technical-documentation/questionnaires.html.

Key Data Items:

Actual, full, and emergency production (establishment must report both actual and full production to be a respondent for the full utilization rates and must report both actual and emergency production to be a respondent for the emergency utilization rates).

Type of Request:

Voluntary

Frequency and Mode of Contact:

The Census Bureau mails the QPC survey letters in March, June, September, and December requesting respondents to complete the survey within approximately 20 business days. An email reminder is sent out one week prior to the due date. A follow-up letter is mailed one week after the due date in an attempt to obtain data from establishments that did not respond. A telephone follow-up is conducted approximately 15 business days after the due date using robocalls, where all establishments with a valid phone number are contacted. Analysts continue to make targeted calls if time allows. A final email is sent 30 business days after the follow-up letter is mailed to any establishments that still have not completed the survey.

Data Collection Unit:

Establishment (data reported based on plant location and activity).

Special Procedures:

N/A

Compilation of Data

Editing:

Standard edits, macro industry review, comparison to sponsor data during the fourth quarter.

Edit tests are run throughout each quarter to determine outliers and reporting errors.  These edit failures are reviewed and resolved by survey analysts.

Industry review is performed at the macro level as delivery dates for the full and emergency utilization rates approach.  Changes in rates from prior quarter to current quarter are flagged and reviewed by survey analysts.  Analysts review and make any necessary corrections, contacting respondents as needed.

For the fourth quarter every year, macro review is performed comparing Census rates and FRB rates.  Analysts review any differences and make any corrections if necessary.

Nonresponse:

Nonresponse is defined as the inability to obtain requested data from an eligible survey unit.  Two types of nonresponse are often distinguished.  Unit nonresponse is the inability to obtain any of the substantive measurements about a unit.  In most cases of unit nonresponse, the Census Bureau was unable to obtain any information from the survey unit after several attempts to elicit a response.  Item nonresponse occurs either when a question is unanswered or the response is unusable.

Nonresponse Adjustment and Imputation:

N/A

Other Macro-level Adjustments:

N/A

Tabulation Unit:

Establishment (data tabbed in one of the 93 industry groups for which data reported by plant).

Estimation:

Full production capacity utilization rates are produced for each of the 93 QPC industry groups using a ratio estimator. Horvitz-Thompson (HT) weighted estimates of actual value of production and estimated value of full production are derived for each industry group using only establishments in the QPC sample that have reported data for both data items. Simple weighted estimates of actual value of production and estimated value of full production are derived by applying the establishment’s sample weight to its respective data values and adding these weighted values across all reporting units in the industry group. The sample weight assigned to each establishment is the inverse of its probability of selection. The full production capacity utilization rate for each industry group is calculated as the ratio of the total weighted actual value of production to the total weighted estimated value of full production. This rate is then multiplied by 100 to yield a percentage. A similar procedure is used to derive the emergency production capacity utilization rates, using actual value of production and estimated value of emergency production. Therefore, the full and emergency production capacity utilization rates for each of the 93 QPC industry groups are derived, respectively, using the following formulas:

Full production capacity utilization rate = [(HT estimate of actual value of production) / (HT estimate of estimated value of full production)] * 100

Emergency production capacity utilization rate = [(HT estimate of actual value of production) / (HT estimate of estimated value of emergency production)] * 100

The average plant hours in operation per week for a given industry group is estimated based on those establishments classified in the industry group reporting plant hours. HT weighted estimates of total plant hours are derived by applying each establishment’s sample weight to its reported plant hours and adding these weighted values across all establishments reporting in the industry group. The average plant hours for a given industry group is then calculated using the following formula:

Average plant hours = (HT estimate of total plant hours) / (sum of weights for all reporting plants)

Comparisons are also made between actual and full production by industry using various checkbox information that is collected in the QPC. This information is collected to determine the primary reasons for changes in full production capacity between current quarter and previous quarter, as well as the primary reasons for actual production being less than full production capacity for the current quarter. Beginning with first-quarter 2013, these data are summarized in the form of weighted proportions for each of the checkbox data items at the 3-digit NAICS level, and historical estimates as far back as first-quarter 2008 are available upon request. Each weighted proportion is the ratio of the weighted number of establishments checking the particular checkbox data item to the weighted number of establishments checking at least one of the checkbox data items for that particular question. Because plants can check more than one of the checkboxes per question, the sum of these respective weighted proportions for a given 3-digit NAICS industry does not necessarily sum to 100 percent. In addition, for any 3-digit NAICS industry with no reported checkbox data for a particular question, both the estimate and the standard error are shown as zero.

Sampling Error:

The sampling error of an estimate based on a sample survey is the difference between the estimate and the result that would be obtained from a complete census conducted under the same survey conditions.  This error occurs because characteristics differ among sampling units in the population and only a subset of the population is measured in a sample survey.  The particular sample used in this survey is one of a large number of samples of the same size that could have been selected using the same sample design.  Because each unit in the sampling frame had a known probability of being selected into the sample, it was possible to estimate the sampling variability of the survey estimates.

Common measures of the variability among these estimates are the sampling variance, the standard error, and the coefficient of variation (CV), which is also referred to as the relative standard error (RSE).  The sampling variance is defined as the squared difference, averaged over all possible samples of the same size and design, between the estimator and its average value. The standard error is the square root of the sampling variance.  The CV expresses the standard error as a percentage of the estimate to which it refers.  For example, an estimate of 200 units that has an estimated standard error of 10 units has an estimated CV of 5 percent.  The sampling variance, standard error, and CV of an estimate can be estimated from the selected sample because the sample was selected using probability sampling.  Note that measures of sampling variability, such as the standard error and CV, are estimated from the sample and are also subject to sampling variability.  It is also important to note that the standard error and CV only measure sampling variability.  They do not measure any systematic biases in the estimates.

The Census Bureau recommends that individuals using these estimates incorporate sampling error information into their analyses, as this could affect the conclusions drawn from the estimates.

The variance estimates for the QPC are calculated using a stratified jackknife method.  Each jackknife estimate is calculated by removing one sample unit, calculating the estimate, and adjusting the estimate upward to account for the missing sample unit.  For estimates of rates, this is done in both the numerator and denominator.  The missing sample unit is placed back into the sample, then another jackknife estimate is calculated after removing the next sample unit.  This process is repeated until all sample units are excluded from one of the jackknife estimates.  The full sample estimate is subtracted from each individual jackknife estimate and final variances are calculated using these differences.  In the QPC publication, estimates are shown along with their corresponding standard errors.

Confidence Interval:

The sample estimate and an estimate of its standard error allow us to construct interval estimates with prescribed confidence that the interval includes the average result of all possible samples with the same size and design.  To illustrate, if all possible samples were surveyed under essentially the same conditions, and an estimate and its standard error were calculated from each sample, then:

1.   Approximately 68 percent of the intervals from one standard error below the estimate to one standard error above the estimate would include the average estimate derived from all possible samples.

2.   Approximately 90 percent of the intervals from 1.645 standard errors below the estimate to 1.645 standard errors above the estimate would include the average estimate derived from all possible samples.

In the example above, the margin of error (MOE) associated with the 90 percent confidence interval is the product of 1.645 and the estimated standard error.

Nonsampling Error:

Nonsampling error encompasses all factors other than sampling error that contribute to the total error associated with an estimate.  This error may also be present in censuses and other non-survey programs.  Nonsampling error arises from many sources: inability to obtain information on all units in the sample; response errors; differences in the interpretation of the questions; mismatches between sampling units and reporting units, requested data and data available or accessible in respondents' records, or with regard to reference periods; mistakes in coding or keying the data obtained; and other errors of collection, response, coverage, and processing.

Although no direct measurement of nonsampling error was obtained, precautionary steps were taken in all phases of the collection, processing, and tabulation of the data in an effort to minimize its influence.  Precise estimation of the magnitude of nonsampling errors would require special experiments or access to independent data and, consequently, the magnitudes are often unavailable.

The Census Bureau recommends that individuals using these estimates factor in this information when assessing their analyses of these data, as nonsampling error could affect the conclusions drawn from the estimates.

Two measures or indicators of nonsampling error are calculated for the QPC.  The first measure is the unit response rate (URR).  The URR is computed as the number of respondents (R) divided by the number of establishments eligible for data collection (E) plus the number of establishments for which eligibility cannot be determined (U).  This rate is then multiplied by 100 to yield the percentage.  Therefore, the formula for calculating the URR is as follows:

URR = [R/(E+U)] * 100

The URR for the QPC is typically 40.0-45.0% from quarter to quarter.

The second measure or indicator of nonsampling error is the coverage rate.  The coverage rate is calculated by dividing the total weighted measure of size (value of shipments) for QPC respondent establishments by the total weighted value of shipments for all active establishments in the QPC sample.  Again, this rate is multiplied by 100 to yield the percentage.  The coverage rate for the QPC is also typically 40.0-45.0% from quarter to quarter.

Benchmarking:

N/A

Disclosure Avoidance:

Disclosure is the release of data that reveals information or permits deduction of information about a particular survey unit through the release of either tables or microdata. Disclosure avoidance is the process used to protect each survey unit’s identity and data from disclosure. Using disclosure avoidance procedures, the Census Bureau modifies or removes the characteristics that put information at risk of disclosure. Although it may appear that a table shows information about a specific survey unit, the Census Bureau has taken steps to disguise or suppress a unit’s data that may be “at risk” of disclosure while making sure the results are still useful.

Disclosure avoidance is done for the QPC using cell suppression. Cell suppression is a disclosure avoidance technique that protects the confidentiality of individual survey units by withholding cell values from release and replacing the cell value with a symbol, usually a “D”. If the suppressed cell value were known, it would allow one to estimate an individual survey unit’s value too closely.

The cells that must be protected are called primary suppressions.

To make sure the cell values of the primary suppressions cannot be closely estimated by using other published cell values, additional cells may also be suppressed. These additional suppressed cells are called complementary suppressions.

The process of suppression does not usually change the higher-level totals. Values for cells that are not suppressed remain unchanged. Before the Census Bureau releases data, computer programs and analysts ensure primary and complementary suppressions have been correctly applied.

The Census Bureau has reviewed the 2020 QPC tables for unauthorized disclosure of confidential information and has approved the disclosure avoidance practices applied (Approval ID: CBDRB-FY20-241, approved April 20, 2020).

For more information on disclosure avoidance practices, see FCSM Statistical Policy Working Paper 22.

History of Survey Program

Prior to 2007, the Census Bureau conducted the annual Plant Capacity Utilization (PCU) Survey. The PCU survey was funded by the Federal Reserve Board (FRB) and the Department of Defense (DOD) and provided estimated fourth-quarter rates of plant capacity utilization. In 2007, the DOD decided they would no longer provide funding for the PCU survey, so the FRB paid for the entire cost of the survey. The PCU survey was converted to a quarterly survey to meet the FRB’s need for more timely data to include in their Industrial Production Index. The data item (emergency production) previously collected for the DOD’s interest, was removed from the QPC survey form at this time.

The new quarterly survey, the Quarterly Survey of Plant Capacity Utilization (QPC), began as a pilot survey in 2007 and has continued as a quarterly survey since then. The purpose of the QPC is to provide estimated quarterly rates of plant capacity utilization for the U.S. manufacturing sector and for the publishing subsector of the information sector. In the summer of 2009, after the OMB approval process began, the DOD again expressed interest in the survey and the collection of emergency production to provide estimated emergency plant capacity utilization rates. The FRB uses the quarterly QPC data to benchmark monthly estimates of capacity output and utilization and to analyze the change in use of capital, capital stocks and inputs related to capacity growth. The DOD uses the QPC data to assess the readiness for meeting the demand for goods under selected national emergency scenarios.

A QPC publication is released approximately 75 days from the completion of each quarter and provides estimated full and emergency plant capacity utilization rates for the 93 industry group levels defined by the survey sponsors. The QPC publication also includes work pattern information and checkbox information about the primary reasons for changes in full production capability from quarter to quarter and the reasons for actual production being less than full production capacity. These are summarized as weighted proportions for each of the checkbox data items.

A sample redesign is conducted for the QPC every five (5) years. The last sample redesign of the QPC was conducted in 2020 for the collection of 2020 data. The next sample redesign is scheduled in 2025 for the collection of 2025 data.