Federal Reserve Economic Data

There’s no hard and fast rule for determining when the U.S. economy has entered a recession, and there’s no one indicator that determines a recession.

The National Bureau of Economic Research (NBER) Business Cycle Dating Committee defines a recession as a significant decline in economic activity spread across the economy and makes that determination by considering numerous indicators of economic activity. They date a recession from the peak of a business cycle through its trough. Most recently, the committee identified February 2020 as the last business cycle peak and April 2020 as the business cycle trough, making this the shortest recession on record—just 2 to 3 months.

While the beginning and ending months of a recession tend to get a lot of attention, the committee also releases the corresponding quarters of the peak and trough. These turning points are primarily based on quarterly averages of the monthly indicators along with prominent quarterly series such as real GDP. In most instances, the turning point quarters match the turning point months. In fact, the months and quarters had all been in alignment since March (Q1) 1954.¹ But for this past recession, the months and quarters of the business cycle turning points do not align.

Again, as far as the monthly turning points go, the most recent business cycle peak was in February 2020 and the trough was in April 2020, implying the only month the economy was clearly in a recessionary state was March 2020. Some time in February, the economy moved from expansion to recession; and some time in April, the economy moved from recession back to expansion.²

For the quarterly turning points, the committee determined that the business cycle peak occurred in the fourth quarter of 2019. This is consistent with the –5.1% decline in real GDP in the first quarter of 2020, but not in alignment with the monthly peak in February 2020. Moreover, the steep decline in employment in March 2020 more than offset the gains in January and February, generating a quarterly decline in employment of over 900,000 jobs (see the FRED Graph above).³

Shaded areas indicate U.S. recessions

Frequent FRED users are familiar with the phrase “Shaded areas indicate U.S. recessions” in the bottom left-hand corner of their FRED graph. You can see this shading below in the FRED graph of the unemployment rate, which also appears on the committee’s web page, with a nod to the St. Louis Fed. The recession shading is a useful visual tool for interpreting economic data. In this case, we can see that the unemployment rate rises during a recession and tends to reach a peak a few months after the recession ends.

The FRED team applies recession shading starting with the month of the peak and ending with the month prior to the trough. This method has most accurately aligned with the turning points in economic data because of the consistency between NBER turning point months and quarters: As noted above, NBER turning point months and quarters have aligned for over a half-century!

Of course, this shading technique is not perfect and things change. The recent discrepancy between the monthly and quarterly turning points may generate some confusion when examining recent data. Rest assured, all the data behind these shaded areas do exist in FRED; and for those users who want to fine-tune their graphs, the series are here.

¹ The committee, which dates back to 1894, has identified 68 business cycle turning points; 13 of the turning months and quarters do not align.
² Currently, daily or weekly economic indicators aren’t robust enough to make this determination. Stock indices, for example, are forward looking; so, turning points in equity prices don’t always correspond with business cycle turning points.
³ This NBER press release discusses the discrepancy in household employment depicted above and nonfarm payroll employment (PAYEMS).

How these graphs were created: For the first graph, search FRED for “Employment Level” and select the series “CE16OV.” The default graph will be a monthly graph of the employment level in thousands of persons. To change units, go to the “Edit Graph” panel: Under “Units,” select “Change, Thousands of Persons.” Next, to add the quarterly employment level, use the “Add Line” tab to search for “Employment Level” and select the series “CE16OV.” Next, under the “Edit Lines” tab, click the “Edit Line 2” button to change the units and frequency. Under “Units,” select “Change, Thousands of Persons.” Then, under “Modify frequency,” select “Quarterly.” Next, under the “Format” tab, select “Off” next to “Recession Shading.” To add marks to the lines, select “Circle” under “Mark type” next to each line. Return to the main graph. Use the date range boxes to set the beginning date to “2018-09-01.”
For the second graph, search FRED for “Unemployment Rate” and select the series “UNRATE.” The default graph will be a monthly graph of the unemployment rate as a percent.

Suggested by Chuck Gascon.

If you’ve been paying rent just about anywhere in the United States, you likely already know that rent has been going up. And the FRED graph above shows exactly that. Average rent in U.S. cities has risen by 85% in just the past 20 years. That’s 30 percentage points above the 55% inflation that’s occurred between then and now (July 2021, at the time of this writing).

Rent growth in the Northeast and South has stayed close to the national average in recent years, while growth in the West has surpassed the national average. The exception is the Midwest, where the regional average has lagged a bit behind the national average. But average rent in all regions still outpaces inflation, with Midwest rent growth remaining the closest (only 7 percentage points above). Since the end of the COVID-19-induced recession, though, inflation has grown faster than rent, potentially a result of the COVID-19-related financial stimulus and rent relief in the form of eviction moratoriums.

The second graph shows that rent growth in specific urban areas pretty closely matches the corresponding regional growth from the first graph. Renting is largely an urban phenomenon, so it shouldn’t be a surprise the biggest cities in these regions accurately reflect the regional trends: Rents in Boston, Detroit, Houston, and Seattle have all grown at rates that very closely track the rates in the Northeast, Midwest, South, and West.

Given these rent increases, are there changes in regional population growth—say, population booms in places with smaller rent growth and busts in places with higher rent growth? Not quite.

Our last FRED graph shows the population growth of two urban areas that are above the U.S. average and two areas that are below it. This dividing line doesn’t separate areas of slow rent growth from areas of fast rent growth, however: Cities like Boston and Houston, with average rent increases, can land in either quadrant. Cities like Seattle are in the upper quadrant despite their high rent increases, and cities like Detroit are in the lower quadrant even with their low rent increases.

FRED allows you to create all kinds of data agglomerations and exercises, so you can examine these and other influential factors, such as vacancy rates, opportunities for building, and income growth.

How these graphs were created:
First graph: Search for “CPI rent primary residence south” and select the series shown. From the “Edit Graph” panel, use the “Add Line” tab to search for and select the other series by replacing “south” with “west,” “northeast,” and “midwest.” Then add the rent CPI for the U.S. and the general CPI series. Change “Units” to “Index (Scale value to 100 for chosen date),” use “2001-01-01″ as the chosen date, and click “Apply to all.” In the “Format” tab, make the last two lines dashed. Adjust the date range to start on 2001-01-01. Second graph: Repeat the above, but for rents in Seattle, Boston, Houston and Detroit. Third graph: Repeat the above, but for populations of the same cities and the U.S. population.

Suggested by Reed Romanko and Christian Zimmermann.

Join us on a road trip of FRED data in search of labor productivity.

The FRED Blog recently compared the increase in labor productivity during the COVID-19-induced recession with labor productivity in past recessions. Today, we use a recently added data set on state-level productivity from the U.S. Bureau of Labor Statistics to compare labor productivity across states.

First, labor productivity is output per hour worked. So, when labor productivity increases, an hour of work yields more output, which means more goods produced or more services delivered with the same amount of effort.

The GeoFRED map above shows the percent growth in labor productivity in Hawaii during 2020. The residents of the very last state to join the Union (August 21, 1959) recorded the fastest growth in labor productivity last year: 8.5%.

How did the other states fare? The GeoFRED map shows Nevadans were not far behind Hawaii residents, as productivity in the Silver State grew 8%. But the blue coloring in the map shows states where productivity growth was negative during 2020. In descending order, an hour of work yielded fewer goods and services than during the previous year for Montanans, Oklahomans, Tennesseans, South Dakotans, and Idahoans.

The regional differences in labor productivity growth likely reflect idiosyncratic state economies. For example, goods manufacturing plays a much larger role in Oklahoma than in Hawaii. In that light, the uneven impact of the COVID-19-induced recession on the nationwide consumption and production of goods and services would have different impacts on state-level output, hours worked, and—ultimately—labor productivity.

How these maps were created: The original post referenced an interactive map from our now discontinued GeoFRED site. The revised post provides a replacement map from FRED’s new mapping tool. To create FRED maps, go to the data series page in question and look for the green “VIEW MAP” button at the top right of the graph. See this post for instructions to edit a FRED map. Only series with a green map button can be mapped.

Suggested by Diego Mendez-Carbajo.