Federal Reserve Economic Data

The FRED® Blog

Posts tagged with: "UNRATE"

View this series on FRED

Losing your job doesn’t always gain you unemployment benefits

Requirements and trends behind unemployment insurance

Being unemployed does not guarantee that you’ll receive benefits from your local unemployment insurance program. Typically, there are eligibility criteria, such as previous work requirements, waiting periods, eligibility periods, and asset tests. These criteria can be stringent, depending on the political choices behind them. The graph above compares the U.S. unemployment rate with the segment of the labor force receiving unemployment insurance benefits. It is very clear that, most of the time, only a minority of the unemployed receive benefits.

The graph below focuses on that segment, showing the proportion of the unemployed that receives insurance benefits. Obviously, there are cyclical variations: At the start of a recession, proportionally more unemployed haven’t yet run out of eligibility. There also appears to be a longer-run trend that has been decreasing the segment of those eligible for benefits.

Update: The insurance claim numbers cover those who get regular state unemployment insurance benefits. There are also those who get benefits under the extended benefit and the emergency unemployment compensation programs, whose proportions tends to be higher during recessions. See this article for an analysis of these details.

How these graphs were created: Search for “unemployment insurance claims” and click on the series. From the “Edit Graph” section, add the “civilian labor force” series and click on “Apply.” Then enter formula a/b/10 (where the 10 makes it a percentage). Then open the “Add Line” tab and search for the unemployment rate; take the monthly, seasonally adjusted series. That’s the first graph. For the second, remove the line you just added, but add that series to the first line and apply formula a/b/c*10.

Suggested by Christian Zimmermann.

View on FRED, series used in this post: CCSA, CLF16OV, UNRATE

Cyclical asymmetry in the labor market

Slow but steady improvements versus sharp declines

Have you ever spent hours on the beach meticulously building the perfect sandcastle, only for a bully to waltz by and kick it down in an instant? A similar phenomenon occurs in the labor market, as even the shortest recessions can undo years of progress made during an expansion. The unemployment rate tends to fall only gradually during economic expansions but rise sharply during recessions. This mismatch of slow declines versus sharp spikes is known as “cyclical asymmetry.”

This cyclical asymmetry of the unemployment rate derives in part from the fact that it takes much longer to create jobs than to destroy them. We can think of the labor market as a market for productive relationships between employers and employees. These relationships are a durable form of capital: They provide long-term economic benefits over time for firms and workers alike. It can take a while to build these relationships but only a short time to terminate them.

Consider the drawn-out process it takes to match employers and employees. First of all, both parties much search for and identify potential matches. Then, employers usually put candidates through a lengthy hiring process before making any offers. And even once the employee is hired, it takes time to establish a working relationship. Hence, job creation tends to be slow, even when the economy is performing well. The unemployment rate does not decline sharply when the economy is hit by a positive disturbance because relationships take time to develop.

Conversely, consider what happens when a recession hits. Letting workers go takes only an instant. In a flexible labor market, firms are often quick to lay off workers to save costs, often letting go workers who have been with the company for years. So, when the economy is hit by a negative disturbance, the unemployment rate tends to spike as firms lay off large numbers of workers.

Cyclical asymmetry also occurs in population dynamics in the form of the “heat wave effect.” Often, mortality rates rise and the population suddenly declines when bad weather hits. Yet, when a streak of good weather hits, we don’t see a corresponding boost in the population, as it takes time to repopulate after a tragedy strikes. In the context of the labor market, a recession is a “heat wave” that leads to sudden job losses and an expansion is a spell of good weather that, over time, creates jobs more steadily.

How this graph was created: Search for “civilian unemployment rate” and pick the seasonally adjusted series from the first result.

Suggested by David Andolfatto and Andrew Spewak.

View on FRED, series used in this post: UNRATE

The full banana of the labor market

An update on the Beveridge curve

Three and a half years ago, we published a blog post about the Beveridge curve featuring the graph above, which shows how job vacancies and unemployment relate to each other. Each dot represents their values at a particular date. Beveridge’s theory is that these two measures don’t form a kinked line along the axes in a scatter plot, but rather a banana shape. This shape occurs because of delays and frictions in the job market: Vacancies and job seekers take time to intersect, as there may be mismatches in terms of job location and qualifications, for example. The graph above doesn’t show the expected full banana because the available sample period just wasn’t long enough. So, we revisit this idea by updating the graph, shown below. The banana, although not very smooth, is now complete.

How this graph was created: Search for “job openings” and add the series to the graph. From the “Edit Graph” section, add the second series by searching for and adding “civilian unemployment.” From the “Format” tab, choose “Scatter” for graph type. To connect the dots, choose a non-zero line width in the settings of the first series, which is where you can also adjust the size of the dots.

Suggested by Charley Kyd and Christian Zimmermann.

View on FRED, series used in this post: JTSJOR, UNRATE


Back to Top