Federal Reserve Economic Data

The FRED Blog uses a variety of graph types (line, area, bar, scatter, and pie) to help tell the story behind the numbers. But, sometimes, data plots reveal features in the data we can’t completely understand.

The FRED graph above shows the unemployment rates for men and women who are at least 25 years old and hold a master’s degree. The data, available since the year 2000, are provided by the U.S. Bureau of Labor Statistics’ Current Population Survey (Household Survey). The metadata about these series tells us the data are not seasonally adjusted, so they’re reported without taking into consideration the recurring ups and downs associated with the seasons. The seesaw shape of the data lines is a telltale sign of those seasonal patterns and that’s what piques our curiosity.

Why does the unemployment rate of women holding master’s degrees regularly spike during the third quarter of the year? We don’t know.

The same pattern is also noticeable among men and women holding a bachelor’s degree, but it isn’t immediately visible among men and women holding associate, professional, or doctoral degrees.

We can’t provide an explanation based on solid evidence because there is no additional data that can help us dive deeper on the topic. For example, although there are data on the percent of employees who are women in multiple industries, such as the education sector, those data are seasonally adjusted and can’t help us illuminate the reason behind the seasonal pattern shown in the graph. As is often the case, more research is needed. Stay tuned to the FRED Blog and we’ll share what we find to help tell the story behind the numbers.

How this graph was created: In FRED, search for “Unemployment Rate – College Graduates – Master’s Degree, 25 years and over, Men.” From the “Edit Graph” panel, use the “Add Line” tab to search for and select “Unemployment Rate – College Graduates – Master’s Degree, 25 years and over, Women.”

Suggested by Lily Levin and Diego Mendez-Carbajo.

In the early stages of the pandemic, stimulus checks and other factors made vehicles a popular large purchase for households. But disruptions in supply chains resulted in a shortage of new vehicles, which are strongly reliant on global supply chains. The shortage in semiconductors played a big role here, given their importance in the production of cars and their components. New vehicle production slowed down as a result, decreasing inventories and increasing prices.

The reopening of the economy and the shortage of new vehicles increased demand for used vehicles, which led to substantial inflation. Indeed, over the course of the pandemic, used car inflation was considered a major driving force of headline inflation.

In January 2022, however, the trend reversed: Used car inflation has started to decline. And, as supply chains are returning to normal, new cars have become more readily available and the excess demand previously directed toward  used cars has shifted back to new cars.

Used cars dealers had been setting their prices or selling their cars at auction for a profit, but they’re now forced to negotiate with buyers, thus pushing prices down. Whether or not these trends continue in the medium term will partially depend on how supply chain disruptions will affect the manufacturing of new vehicles.

How this graph was created: Search FRED for “CPI Used Vehicles” and select “Consumer Price Index for All Urban Consumers: Used Vehicles and Trucks in U.S. City Average.” From the “Edit Graph” panel, change the units for both lines to “Percent change from a year ago.”

Suggested by Ana Maria Santacreu and Jesse LaBelle.

One of the most watched U.S. economic indicators is the growth of real gross domestic product (GDP). A similar but lesser-known economic indicator has also been in the news lately—real gross domestic income (GDI). According to the Bureau of Economic Analysis, both real GDP and real GDI measure the real output of the U.S. economy. Real GDP measures the value of goods, while real GDI measures the income of employees and corporations. In theory, the growth rates of real GDP and real GDI should be equal.

Lately, this hasn’t been the case. The FRED graph above displays the compounded annual rate of change of real GDP and real GDI over the past 10 years. As the graph shows, growth in GDP and GDI have both slowed recently and have also diverged, with real GDP growth turning negative and real GDI growth remaining positive.

What’s going on here? Although there’s no consensus, a few possible explanations include

• measurement error
• missing data
• sampling errors
• non-sampling errors such as survey nonresponse
• business cycles

Although the statistical discrepancy tends to change quarter to quarter, some scholars have found predictability in this change. When GDP and GDI data are revised, the revisions tend to be smaller for GDI, which may make it a more-accurate indicator than GDP. See Owyang (2016) for more details. For example, GDI growth is currently higher than GDP growth; in the past when this has occurred, GDP has typically been revised up.

There’s room for debate on whether GDP or GDI should be the primary economic indicator for determining the health of the economy. GDP is released in a more timely manner, which may explain why it garners more attention. However, it’s clear that both measures of economic output offer valuable information on the health of the economy.

How this graph was created: Search FRED for “real gross domestic income” and select the series “A261RX1Q020SBEA.” The default graph will be a quarterly graph of the gross domestic income in terms of billions of chained 2012 dollars. Use the “Edit Graph” button to open the editing box: Here, change the series into a rate of change and add the real gross domestic product series. In the “Units” dropdown menu, change it to “Compounded Annual Rate of Change.” Next, use the “Add Line” tab to “Create user-defined line.” Search for “real gross domestic product” or equivalently the series name “GDPC1.” Click “Add series” and make sure the units are also in compounded annual rate of change. Return to the main graph. Use the date range boxes to set the beginning date to “2012-04-01.”

Suggested by Charles Gascon and Cassie Marks.