Federal Reserve Economic Data

The FRED® Blog

The vanishing negative saving rate

One data series that’s been subject to frequent revision, and in the past frequent inclusion in the news, is the personal saving rate. In October 2007, our Liber8 publication (a newsletter for librarians) included an explanation of the negative U.S. saving rate.* To make a long story short, that article and others from that time noted that it generally isn’t wise to spend more than your income. However, if you look at the current saving rate, you see the graph never goes below zero. It even seems to stay above 2%. What’s up with that? Well, revisions to the data.

Below is a graph of changes in the data, comparing the current data with the data available in September 2007.

How this graph was created: Personal saving rate data were most recently released on March 28, 2014. To see previous releases, select the ALFRED link, which is to the left of the graph. Edit data series 1 by changing the “as-of date” to August 31, 2007. Then change the graph from a bar graph to a line graph.

Suggested by Katrina Stierholz.

* In our defense, there were other Fed publications that wrote about the negative saving rate as well.

View on FRED, series used in this post: PSAVERT

Gender labor force participation gaps

In a recent article in the Regional Economist, Li Li and I plotted gender gaps in labor markets of G7 countries in 1991, 2001, and 2011. We focused in particular on two gender gaps: labor force participation (the difference in labor force participation rates for men and women) and unemployment (the difference between the unemployment rate for men and women).

While we used data from the World Bank and OECD, we could have just as easily transformed FRED data series to obtain these gaps. In the graph, we show the gender labor force participation rate gap in the United States and the G7 country with the largest gender labor force participation gap in the most recent data: Italy.

In the United States, the labor force participation rate for men has been trending down since World War II, while the labor force participation rate for women has been trending up—that is, until the Great Recession, when it began trending down in parallel with the rate for men. The gap has shrunk to about 10 to 12 percent. In Italy, the trend of the gap is very similar, but it’s shifted to the right and to date the gap is still almost twice as large as that in the United States.

How this graph was created: After finding the first series (men, United States) and the second series (women, United States), create your own data transformation using the formula a-b. After finding the same two series for Italy, apply the same transformation.

Suggested by Silvio Contessi

View on FRED, series used in this post: ITALFPMNA, ITALFPWNA, LNS11300001, LNS11300002

The Taylor Rule

This graph shows in blue the Taylor Rule, which is a simple formula that John Taylor devised to guide policymakers. It calculates what the federal funds rate should be, as a function of the output gap and current inflation. Here, we measure the output gap as the difference between potential output (published by the Congressional Budget Office) and real GDP. Inflation is measured by changes in the CPI, and we use a target inflation rate of 2%. We also assume a steady-state real interest rate of 2%. These are a lot of assumptions, and you are welcome to change them on the graph by playing around with the formula to see how the Taylor Rule matches up with the effective federal funds rate. To read up on the Taylor Rule, see the original article or an article by former St. Louis Fed president William Poole.

How this graph was created: To create a new series from several series, first add the series by modifying the existing series in the “Graph” tab. Once you have assembled them all, expand the series section in the same tab and “create your own transformation.” Finally, as the axis legend has become unwieldy, remove it by checking off the mark in the graph tab.

Suggested by: Christian Zimmermann

Update: A previous version did not multiply the output gap by 100.

View on FRED, series used in this post: FEDFUNDS, GDPC1, GDPDEF, GDPPOT


Back to Top