The Employment Cost Index (ECI) is an index that measures change in employers’ costs for employee wages and benefits. It is designed to measure the aggregate cost from one period to the next. Three parts are needed to calculate employers’ costs for employer-provided benefits:
Collection of benefit costs using these items is referred to as “rate and usage” collection.
Usage is the utilization of a specific benefit by employees. While all workers in the job are offered the benefit, not all will have met eligibility criteria. Individuals may choose not to participate in a benefit even after meeting eligibility criteria. Participation below 100 percent is more common when there is an employee cost requirement, such as matching contributions for a retirement plan or monthly health insurance premiums. Employee benefit usage is collected to better reflect the employer cost for the workers in the sampled job.
How usage is measured and collected depends upon the benefit itself. For instance, overtime pay is a time-based benefit and usage is defined as the number of overtime hours worked by the employees. Health care is a plan-based benefit and usage is the distribution of employees' enrollment among offered plans.
When detailed rate and usage data are not available, other options for capturing employer costs exist, such as collecting the employers’ expenditure for benefits along with the number of workers offered the benefits. Expenditures reflect employers’ cost and the employees’ usage; however, they are often for a group of workers larger than sampled jobs. Rate and usage data is preferred to expenditure data because it often provides more occupational detail.
Employers’ costs for providing benefits will change as the rate or usage changes. There are incidental shifts in usage as workers change or experience life events that drive benefit participation. The ECI is not designed to measure this level of fluctuation in employers’ expenditures. The ECI is based on observing the changes in the cost of employing a fixed set of labor inputs, primarily accomplished by fixed occupational and industry weights.
There is not a conceptual requirement for holding usage constant in a fixed employment weight labor cost index. However, there is a respondent burden reduction by holding usage constant after the benefit cost is initiated. This is an important consideration in a voluntary survey with declining response rates.
Holding usage constant is a collection efficiency that permits the field economist to focus on collecting changes in the benefit rate and the benefit provisions. New rates can be applied to the previously collected usage to calculate a new employer cost that reflects the change in rate. As noted, expenditures will reflect changes in usage. Expenditure data is updated when the underlying rate for the benefit changes, ensuring changes are not driven solely by incidental worker preferences.
The ECI collects changes in benefit packages offered by employers to sampled employees. When benefits are added or removed, these changes are reflected. Changes to the specific details of these benefits such as the number of days offered, the cost sharing ratio of premiums, or the amount of coverage provided are identified and benefit costs are updated.
The ECI includes induced changes in usage. An induced change in usage is a change in usage caused by a change in employees' decisions due to changes to a benefit. For example, employees may choose to opt out of life insurance if the employer began requiring employees pay a portion of the premium. When the benefit provisions or cost sharing change, all portions needed to calculate the cost are collected again regardless of how the original data were collected.
The rotation of sample units brings shifts in labor market conditions into the index. While usage is only collected at the time the benefit cost is initiated, one third of the private sample is replenished each year. Using overtime as an example, the hours collected at initiation for a job would change only if there was a change to the plan or work schedule at that specific establishment. This seldom happens so the overtime hours (usage) are unchanged, and the index is largely driven by changes to earnings. However, new sample members are initiated each year and the newly collected overtime hours reflect current economic conditions.
 BLS publishes response rates for household and establishment-based surveys, see charts 2 and 3 for information overall unit and initiation response rates, respectively within www.bls.gov/osmr/response-rates/.
 With the December 2021 reference period 4 private industry samples are used in estimation, such that 25 percent of the private sample is rotated annually.