Federal Reserve Economic Data

The FRED® Blog

Gasoline prices and consumer expenditures

Part of the “My favorite FRED graph” guest post series.

The FRED graph above shows gasoline prices and real personal consumption expenditures on motor vehicle fuels, lubricants, and fluids between 2006 and 2015.

I use this graph when illustrating how changes in prices influence the quantity demanded of a good or service. The inverse relationship between these two variables is a foundational concept in introductory economics courses. This relationship is also used in econ courses to calculate the price elasticity of demand.

But instead of using made-up figures or imaginary scenarios to illustrate the price elasticity of demand, I direct my students to use real data from FRED. My co-author and I published an article in Journal of Economic Education presenting a step-by-step process for instructors to use this type of FRED data in their classrooms: My students download the data behind this graph or hover over it to write down price and quantity index values; then, they use conventional formulas to calculate consumers’ responsiveness to price changes between any two dates.

And this exercise has a bonus teachable moment. In a couple of occasions, the elasticity of demand coefficients are positive, thus providing an opportunity to discuss the “non-price determinants” of demand. For example, the positive elasticity coefficient between 2008 and 2009 provides an opportunity to discuss the Great Recession and how it may have affected real personal consumption expenditures on motor vehicle fuels, lubricants, and fluids. Data like these help me bring economics to life.

A note about the data in the graph: This line graph shows two data series from two different sources: The blue line shows conventional gasoline prices, measured in dollars per gallon, as reported by the Energy Information Administration. The red line shows the quantity of real personal consumption expenditures on motor vehicle fuels, lubricants, and fluids, measured using an index, as reported by the U.S. Bureau of Economic Analysis.

How this graph was created: Search for and select “Conventional Gasoline Prices: U.S. Gulf Coast, Regular.” From the “Edit Graph” panel, use the “Add Line” tab to search for and select “Real personal consumption expenditures: Nondurable goods: Motor vehicle fuels, lubricants, and fluids (chain-type quantity index).” To change the date range, adjust the start and end dates above the graph.

Suggested by Carlos J. Asarta.

A data story of a segmented labor market

Although the FRED Blog is committed to the long haul, we’ve discussed quitting before: that is, data on voluntary separations from employment. For example, earlier this year, a post covered quits for currently employed workers during economic downturns.

Conventional wisdom says that quits decrease during recessions and increase during expansions. But because there was hardly anything conventional about the COVID-19-induced economic shock, we take a look at the recent evolution of voluntary separations. And from the graph above, we see that the overall level of quits has reached new heights.

The graph plots data from the Job Openings and Labor Turnover Survey from the U.S. Bureau of Labor Statistics (BLS), showing the percentage rate of quits among all nonfarm employees (in black), private-sector employees (in blue), and government employees (in red). Because private-sector employees greatly outnumber government employees, the all-employees quit rate most closely reflects the labor market experiences of that private-sector group. Yet, regardless of the relative size of each section of the labor force, private employees quit their jobs three times more frequently than government employees do.

The data show some milestones: In August 2021, the overall quit rate hit at an all-time high of 2.9%. Private-sector employees have been quitting their jobs at unprecedented rates since April 2021, reaching 3.3% in August 2021. In contrast, the record-high quit rate among government employees is 1.1%, reported in both July and October of 2020. It has declined since then and seems to be returning to its very stable recent average of 0.8%.

These developments provide additional evidence that the private and public groups occupy different segments of the labor market. The January 2020 BLS survey on employee tenure reports government workers keep their jobs almost twice as long as private-sector workers do. Researchers at the BLS point out the former are, on average, older than the latter. No doubt the incentives to remain on the job significantly change with age, but when overall labor conditions change dramatically, perhaps the sunk cost fallacy (i.e., staying on a course of action due to a cost that has already been incurred and cannot be recovered) is unduly affecting some workers’ choices.

How this graph was created: Search for and select “Quits: Total Nonfarm,” using the “Monthly, Rate, Seasonally Adjusted” series. From the “Edit Graph” panel, use the “Add Line” tab to search for and add “Quits: Total Private” and “Quits: Government” to the graph. Remember to select the data frequency and units listed above.

Suggested by George Fortier and Diego Mendez-Carbajo.

View on FRED, series used in this post: JTS1000QUR, JTS9000QUR, JTSQUR

Since the pandemic and its recession, service sector earnings have risen relative to the goods sector

The FRED Blog has discussed the demand and supply sides of the ongoing labor market shortages. Today, we examine the potential impact those shortages may have on labor compensation. That is, does the current level of job openings have an effect on earnings?

The FRED graph above plots data from the monthly Establishment Survey, which comes from the U.S. Bureau of Labor Statistics’ Current Employment Statistics program. It shows the average weekly earnings of all employees on payrolls in the service industry relative to those of employees in the goods-producing industry. The data are seasonally adjusted, so changes caused by seasonal factors such as back-to-school retail sales won’t distort the comparison.

The data (from March 2006 through September 2021) show that, until March 2020, for each dollar earned by a goods-producing employee, a service-providing employee earned approximately 78 cents. The COVID-19 pandemic, the recession it induced, and its aftermath changed that proportion. As of September 2021, for each dollar earned by a goods-producing employee, a service-providing employee earned approximately 81 cents. That’s almost a 4% increase.

Although we can’t put forward any single reason for the change in relative earnings, recent research by Sang Yoon Lee, Minsung Park, and Yongseok Shin looks at the unequal effects of the COVID-19-induced economic shock on the labor market. The authors find significant differences in employment across industries, occupations, and place of residence.

How this graph was created: Search for and select “Average Weekly Earnings of All Employees, Private Service-Providing.” From the “Edit Graph” panel, use the “Edit Line 1” tab to customize the data by searching for and selecting “Average Weekly Earnings of All Employees, Goods-Producing.” Last, create a custom formula to combine the series by typing in “a/b” and clicking “Apply.”

Suggested by Diego Mendez-Carbajo.

View on FRED, series used in this post: CES0600000011, CES0800000011