OLMIS

by Ken Lux, Gary Sincick
Published Oct-1-2000

What is it? Firms and Reporting Units Average Pay per Worker Covered vs. Noncovered Time Series Analysis Non-economic Changes Changes in Industry Classification Changes in UI Coverage

Oregon covered employment and payroll information is one of the most formidable tools available for analyzing Oregon's statewide and regional economies. Like any powerful tool, of course, it requires care in its use and some understanding of its operation. To get the most out of the data, it is helpful to have a solid understanding of how it is gathered and compiled.

Unlike data that come from sample surveys, Oregon covered employment and payroll information is based on tax reports submitted quarterly by employers subject to Unemployment Insurance (UI) law and by the program of Unemployment Compensation for Federal Employees (UCFE). Thus, 'covered' employment and payroll refers to workers and wages that are covered by unemployment insurance.

Employment is reported by firms for each month using the payroll period that includes the 12th day of each month. Wages are reported for the entire quarter. For example, if a company pays its employees twice a month, the employment reported would be only the number of employees who worked during the first payroll period of each month. The total wages reported would be the sum of both pay periods for each of the three months in the quarter.

Thus, the nature of the way that the figures are reported means that employment is a point-in-time count, rather than a measure of full-time equivalency. For example, if a company's payroll is semi-monthly and the employer has five employees during the month, but one started work on the 20th of the month, only four employees would be included in that month's employment tally.

Summary level employment and payroll data are based on ownership (e.g. private, federal, state, etc.), county, and industry classification codes (currently SIC, soon to be NAICS) assigned by the Employment Department. The most detailed industry data are available at the statewide level, with county data available at a higher level of aggregation. State Employment Department law regarding confidentiality of information collected through the Unemployment Insurance tax reports does not allow the publication of employment, wage, or any other data that could be identified with an individual employer.

Employment and payroll data are ordinarily classified on the basis of the primary business activity and physical location of the employer in Oregon. Employers operating at more than one location in Oregon are required to report their employment and payroll by each location, and each branch is classified according to its primary activity. This makes it possible to identify and report employment and wages in the area in which the work takes place and where the economic impact of the jobs is the greatest.

Data on the number of "reporting units" for a given industry and geographic area can be loosely thought of as the number of firms. There is an important difference between "firms" and "reporting units", however. A reporting unit is defined as either a single-location company or an individual physical location of a multiple-location company (i.e., one store in a chain of stores). A "firm", on the other hand, may encompass multiple reporting units (i.e., the entire chain).

Grouping employment by reporting unit is often more useful than grouping it by firm. At the firm level, all employment and payroll for the firm is assigned to a single industry and location. Some operations of the firm may fall outside its assigned location. This information is captured, however, when breaking the data down by reporting units. Thus, the distinction between firms and reporting units becomes particularly important when analyzing data for smaller geographic areas, such as counties.

Once having gathered and compiled employment and payroll data by industry, it is simply a matter of dividing payroll by employment to calculate the average pay per worker. For example, to calculate Oregon's annual average pay per worker in the manufacturing industry for 1998, we take the total annual payroll of $9,522,802,688 and divide it by the annual average employment of 245,656 to get $38,765.

These figures are often used to make rough comparisons of wages across different industries and geographic areas. When making such comparisons, however, it is important to keep in mind that the average annual pay per worker in a particular industry is going to be affected not only by the wage level in that industry, but also by the prevalence of part-time employment.

Consider a hypothetical example of two industries, each with 100 employees, all of whom earn an hourly wage of $10 per hour. The only difference between the two industries is that in one, all employees work 40 hours per week, and in the other all employees work 20 hours per week. Even though employment is the same (as measured by covered employment data) and wages are the same, one industry will show twice the level of annual average pay per worker as the other. This difference is extremely important to take into consideration when comparing two industries in which the level of part-time employment differs considerably, such as manufacturing and retail trade.

Not all employed persons in Oregon are included in the tallies of covered employment and payroll. That is, not everyone who works in Oregon is covered by the UI program. The Oregon Employment Department does prepare estimates of total (covered and non-covered) employment. Since 1980, the ratio of covered employment to total employment, expressed as a percentage, has varied between a low of 78.6 % (in 1983) and a high of 95.2% (in 1997). During those years, the covered employment ratio averaged 87.9%. The complete definition of covered versus non-covered employment is contained in Employment Department law, Chapter 657 ORS, sections 657.010 through 657.097

There are a number of specific groups that are, by law, excluded from UI coverage and are thus not counted in the covered employment and payroll statistics program. Employment and payroll data on interstate railroad workers, who are covered under a separate unemployment insurance law administered by the Railroad Retirement Board, are not included. Also excluded are the following kinds of employment:

1. Self-employment.
2. Agricultural labor performed for a farm with a quarterly payroll of less than $20,000 or not employing at least 10 persons in each of 20 separate weeks during any calendar year.
3. Domestic service in a home, sorority, or fraternity, providing the quarterly payroll at no time exceeds $1,000.
4. Casual labor not in the course of an employer's trade or business.
5. Service performed as an officer or member of the crew of any American vessel primarily engaged in interstate, foreign, or high seas navigation, which does not maintain an office within Oregon from which the operations of the vessel are regularly managed and controlled, and service performed on any vessel of foreign registry. (Officers and crews of vessels engaged in inland navigation on the Willamette and Columbia Rivers are covered.)
6. Service performed by a person in the employ of a son, daughter, or spouse, and service performed by a child under the age of 18 in the employ of his father or mother.
7. Service performed for religious organizations by members of the clergy. (Note: effective with the fourth quarter, 1996, tax reports, this exception is no longer applicable.)
8. Service performed by certain part-time, irregular, and emergency employees of state or local government.
9. Service performed by elected officials.
10. Service by an appointed policymaking official of state or local government provided he or she works less than 8 hours a week.
11. Service performed by an individual in the delivery of newspapers or shopping news.
12. Service performed by a real estate broker, real estate salesman, real estate agent, insurance agent, insurance solicitor, or securities salesperson to the extent that compensation is solely by commission.
13. Service performed by an individual or partnership in the distribution of petroleum products with remuneration for service primarily consisting of the difference between the amount the individual pays or is obligated to pay for the petroleum products and the amount the individual receives.
14. Commission sales of home improvements and in-home sales of consumer goods.

Probably the most valuable use of the covered employment and payroll information is the ability it gives the user to observe the year-to-year trends of employment in Oregon. However, as previously noted, there are certain restrictions that make strict year-to-year comparisons misleading or outright impossible. Technical changes in the Unemployment Insurance program, changes in the industrial classification system, or our ongoing review of assigned industry and location codes can sometimes cause the appearance of sudden changes in employment and payroll tables, either as gains or losses.

Such changes do not accurately reflect changes in the economic structure of Oregon's economy and as such may limit the legitimacy of year-to-year comparisons of data. This should be kept in mind when analyzing employment and payroll trends over time. The three major factors to consider when making year-to-year comparisons are examined in detail in the following sections.

Non-economic changes are defined as those that are technical in nature and not caused by immediate and identifiable changes in the state's economy. The primary causes of non-economic changes in employment and payrolls are industry or county reclassifications. Summary tables of non-economic code changes are made available each year in the Oregon Covered Employment and Payrolls publication.

An annual refiling survey is the major source of non-economic code changes. The focus of that survey is to review ownership, county, and industry codes assigned to describe each reporting unit. One-third of all Oregon employers are surveyed each year with the changes incorporated in the publication the following year. Note that since these code changes become effective in January of each year, a visual scan of the data can often reveal their impact.

For example, a non-economic code change resulted in an apparent drop of over 400 in cutlery manufacturing employment (SIC 3421) between December 1997 and January 1998. Graph 1 below shows how the code change interrupts the series. It is also a reasonable assumption, based on graphic depiction of the series, that a non-economic code change was responsible for the relatively large increase in employment in January 1995.

As Graph 1 shows, relatively large non-economic code changes can be spotted easily in graphs of the monthly employment data. However, it is very important to be aware of the fact that even an apparently consistent data series may contain significant non-economic code changes that may have a crucial effect on the interpretation of the data. To be certain of the impact of these code changes, refer to the tables of non-economic code changes contained in the Covered Employment and Payrolls annual publication.

Changes in the level of detail reported by some employers is the other primary cause of employment and/or wage variations that are non-economic in nature. Beginning in 1989, the Bureau of Labor Statistics, in conjunction with the states, initiated the Business Establishment List (BEL) project to obtain a more accurate representation of employment at the local level. As a result, employers that operate in more than one location and have at least ten employees outside their primary location are required to report their employment and payroll by separate location. When the county or SIC code of the individual location is different from that listed on the master account, the effect on the data is the same as the industry reclassifications described above.

Several revisions of the SIC Manual have occurred. Thus, employment and payroll data defined by SIC code is not always strictly comparable from year-to-year for those industries where changes have been introduced. The following is a list of classification changes with the year such changes were made effective. Details on specific industry changes can be found in the SIC manuals. Unless otherwise noted, changes became effective on January 1 of the year listed.

1943	Non-manufacturing industries were classified in accordance with the Social Security Board Industrial Classification Code, 1942 edition.
1947	The 1945 edition of the SIC manual came into use for classifying manufacturing industries.
1958	Data was classified according to the 1957 edition of the SIC Manual. This revision provided an industrial classification system which took into account the significant technological and structural changes that occurred in the American economy since the development of the previous SIC Manuals.
1963	A supplement to the 1957 Manual had a slight effect on the comparability of data in one or two industries between 1963 and 1964. 1968 Employment and payrolls were classified in accordance with the 1967 edition of the SIC manual. The effect of this revision was of little consequence as far as comparability of data between 1967 and 1968 was concerned.
1975	The 1972 edition made sweeping changes, not only in the composition of the classifications, but also in the basic structure of the Manual. Revisions included changing government from an industry type to an ownership classification, and improvements in industry detail.
1988	The changes introduced in the 1987 edition of the SIC Manual were introduced to improve industry detail, coverage and definitions. These changes were necessary due to the structural, technological and institutional changes in the American economy, for example the rapid expansion of the services and high technology sector.
1997	The introduction of a new coding system, the North American Industry Classification System (NAICS) began with the publication of the NAICS manual in 1998. This new six digit classification system will completely replace the existing four digit SIC system. However, due to the length of time required to re-classify all establishments in the US economy, these changes will not take effect until the year 2000 employment statistics are published, at the earliest.

Revisions in the scope of Unemployment Insurance law coverage have had an impact on the comparability of employment and wage data from year to year. The following is a list of coverage changes with the year such changes were made effective. Unless noted, changes became effective on January 1 of that year.

1936	Oregon covered employment data included only employment and payrolls of private industry employers hiring four or more persons.
1956	Coverage was extended to private firms hiring two or more employees and to employees of federal government installations.
1958	State government workers were added to covered employment totals.
1960	Coverage was extended to all private industry employers with one or more employees.
1972	Employees of nonprofit institutions, other than religious and primary and secondary schools; share-of-the-catch fishermen; employees of commercial plants engaged in handling, grading, shipping, etc., of farm commodities; and faculty members of the state of Oregon higher education were covered.
1974	Coverage was extended to include local government.
1978	Some farm workers and domestic employees were added as well as coverage of primary and secondary private schools. The coverage of agricultural and domestic employment was limited by size of payroll and/or number of employees. At the same time, the exemption of commission sales was expanded to exclude all in-home sales.
1987	The Oregon Legislature introduced two minor changes. Food product demonstrators not employed by the product manufacturer, distributor or retailer were excluded. Also excluded was transportation performed by motor vehicle for a certified common carrier by any person that leases their equipment to a certified common carrier and that personally operates, furnishes and maintains the equipment.
1989	All churches and religious organizations must provide unemployment insurance for all lay employees. Members of the clergy remain exempt.
1996	Members of the clergy and religious organizations, formerly exempt, are now covered beginning with fourth quarter, 1996, Quarterly Tax Reports.

We realize that these changes present difficulties in the analysis of time series data. If, in your analysis, something appears unusual, or you simply wish to confirm your conclusions, please feel free to contact the Research and Analysis section of the Oregon Employment Department. We will be glad to assist where we can.