Is there a relationship between the number of people employed in Wyoming and how much money they make? In other words, if there is a five percent increase in the number of new jobs created would there be a corresponding five percent increase in earnings over the same period of time? What if there was a five percent increase in jobs but a five percent decrease in wages? The total Nonagricultural Wage and Salary Employment estimate for March rose 2.4 percent from a year ago, but did the wages increase or decrease? What, if any, information would these correlations tell us about Wyoming's employment structure? This article will attempt to find answers to these questions.
Each quarter, employers throughout the State who have employees covered by unemployment insurance, submit a report that lists the number of people working each month of the quarter and the total wages for the quarter. Every employer is assigned an industry code based on their major work activity and classified as a private or government organization. The data are then compiled and grouped according to these classifications. For the analysis in this article, the employment and wages in first quarter for 1989-1994 will be used. Table 1 shows the employment and wages for 1992-1994 and the percent changes from the previous year for the major industry groups.
The relationship between employment and wages seems to vary according to the industry. In 1994, for example: Services employment increased about 5.8% and the wages increased about the same amount (4.8%); Construction employment jumped up 15.2 percent while the wages barely changed (1.7%); Transportation & Public Utilities (TPU) had a 3.9 percent increase in wages however the employment decreased 0.9 percent; Finance, Insurance, & Real Estate (FIRE) had a slight increase in employment (2.7%) but the wages increased substantially (26.0%). Each industry appears to behave independent of the other industries.
Will the relationships in each industry's employment and wages be the same every year? The logical answer is no. It may be typical for some employers to have existing employees work longer hours before hiring additional employees. Some industries tend to hire seasonal or temporary employees and pay these people wages that are significantly lower than permanent employees. Perhaps a new firm moves to Wyoming and pays wages that are considerably higher or lower than what is normal for an industry. These are examples of circumstances which could cause the employment and wage relationship to vary each year. Looking at Table 1 and comparing 1993 to 1994, it is evident the relationship is not necessarily the same: in 1993, FIRE increased employment 8.8 percent but the wages only increased 4.1 percent. This is a dramatic difference from what happened in '94 in this industry. However, the Services industry seemed to exhibit relatively the same proportional relationship from 1993 to '94.
The correlation number listed in Table 2 shows the relationship between employment and wages for each industry. This number is a statistic used to determine the association between variables. Its value ranges between zero and one. The closer the correlation number is to one, the stronger the relationship between employment and wages. An employment level change in Total, Retail Trade, Services, or Government should exhibit the same relative change in wages from year to year since the correlation number is close to one for these industries. An employment level change for Mining, TPU, or Wholesale Trade, on the other hand, would have little or no corresponding change in the wages from one year to another.
Because data are collected for all employees in Wyoming covered by Unemployment Insurance (approximately 18,000 employing units in 1994), the compilation of this data into summary statistics takes several months. For instance, the first quarter of 1995 is over but the employment and wage data will not be available until August or September.
On the other hand the Nonagricultural Wage and Salary Employment figures are published every month in Wyoming Labor Force Trends. These numbers are obtained by surveying employers (about 3,600 employing units) each month. Figures from this survey are available sooner because there is less data to compile. March employment estimates are printed in this issue, therefore first quarter numbers can be obtained by taking an average for the three months. This average (in thousands) is contained in Table 2 along with the percentage change from '94. As its name indicates, the survey does not include agriculture employment. It is also designed differently because it includes some employment not covered by Unemployment Insurance.
Table 2 lists the employment in first quarter '94 for covered (Unemployment Insurance) and Nonagricultural employment. In comparing, the data are identical for most industries with a few notable exceptions. These exceptions are due to non-covered employment in TPU, Services, and Government. The TPU nonagricultural figure includes railroad transportation. The railroads have their own system to compensate employees if they are unemployed. Nonagricultural Services employment includes church and other employment from nonprofit organizations. It also includes working students and working spouses of students on the payrolls of private colleges and universities who are not covered by Unemployment Insurance. Government employment includes the same definition at public colleges and universities as well as interns and trainees at hospitals, elected officials, and paid volunteer fire fighters. Even though the design is different for the two systems the percentage changes for each industry is almost identical.
The nonagricultural estimates can be used to gauge the economy before the covered employment numbers are available. Total employment is up 2.7 percent from first quarter 1994. Significant increases occurred in Construction (9.1%), Manufacturing (5.2%), Wholesale Trade (4.4%), Retail Trade (3.9%) and Services (3.6%). The correlation number between the nonagricultural employment estimates and the covered wages is also in Table 2. The correlation number is very comparable between the two systems. Since employment estimates are available for first quarter, it is possible to predict the wages for each industry. In doing this however, attention should be given to the correlation number in producing wage estimates. Those industries with higher correlation numbers should have their actual wages closer to the predicted values. For this reason, the predicted wages are not listed for industries whose correlation number is less than 0.5. The predicted wages and their predicted percent changes are included in Table 2. FIRE is the only industry predicted to have a decrease in wages (5.0%), while Retail Trade is predicted to increase the most (5.1%). When the covered employment and wages are released later this year for first quarter, it will be interesting to see how the predicted and actual wages compare.
At the time he wrote this article, Brett Judd was a Statistician specializing in Nonagricultural Wage and Salary Employment/Current Employment Statistics (CES).