Federal Reserve Economic Data

The FRED® Blog

Live by the barrel, die by the barrel

Connections between oil production, oil dependency, and economic growth

In every introductory macroeconomics course, oil is used as the classic example of a negative price shock. Professors tend to discuss the 1973 oil price shock triggered by the Arab-Israeli conflict and the 1979 oil price shock caused by the Iranian Revolution as reasons for rising inflation and falling global output—connecting these shocks to models about investment and aggregate supply and demand. More recent literature, including this presentation by St. Louis Fed President James Bullard, indicates that oil prices can sometimes be interpreted as a proxy for demand. But what’s the impact of oil supply for the consumers in oil-producing countries? We can use FRED to plot crude oil production versus GDP growth in oil-producing countries to get at least a first idea of just how oil-dependent a country might be.

For the United States, the relative importance of oil to industrial production (which is now less than 20% of the economy) is typically between 7% and 15%. Thus, in the graph above, the correlation between oil production and GDP growth per capita is practically negligible. In fact, the correlation is slightly negative. It’s unlikely that changes in oil production have much of an effect on aggregate economic activity.

But the relationship between oil production and GDP growth per capita is much stronger for countries that have more oil-dependent economies. For example, the correlation coefficient for this measure is 0.51 for the United Arab Emirates, 0.76 for Iran, and 0.93 for Iraq. (The closer this coefficient is to 1.0, the stronger the positive correlation.) The scatter plot below indicates the strength of this positive relationship. For these countries, aggregate well-being could be largely influenced by how much oil the country produces—which is why economic diversification is key to building a national economy less susceptible to oil or other shocks.

How these graphs were created: For the first graph, search for and select “constant GDP per capital United States” and click “Add to Graph.” From the “Edit Graph” panel, use the “Add a Line” feature to search for and select “industrial production crude oil”; change the units to “percent change from year ago” in the “Units” dropdown menu and click “Copy to All.” In the “Format” tab, change the line type to “Scatter Plot.” For the second graph, search for and select “constant GDP per capita United Arab Emirates” and click “Add to Graph.” From the “Edit Graph” panel, use the “Add a Line” feature to search for and select “crude oil production United Arab Emirates.” Repeat this process for each individual country. Change the units to “percent change from year ago” in the “Units” dropdown menu and click “Copy to All.” Change the line graph to a scatter plot by using the “Format” tab and changing “Graph type” entry to “Scatter” and pick different colors as needed.

Suggested by Darren Chang and Christian Zimmermann.

View on FRED, series used in this post: IPG211111CN, NYGDPPCAPKDUSA

Switzerland’s mountainous monetary base

More Swiss uniqueness on their national holiday

Today is the Swiss national holiday. In the past, we’ve taken this opportunity to discuss some unique (i.e., weird) feature of the Swiss economy. This time we use FRED to compare the Swiss monetary base with the U.S. monetary base. To make them comparable, we divide each by its country’s nominal GDP. We see that the general patterns are similar, with a sudden increase in 2008. While the U.S. monetary base has started to go back down (it’s lost a quarter since its high point), there’s nothing that shows any tendency to return to the long-run trend. Indeed, Switzerland is still working with extremely low (even negative) interest rates.

But let’s talk about the stark difference shown in the graph. This statistic for Switzerland is dramatically higher than it is for the U.S.: The Swiss monetary base is now worth over three years of its GDP, while the U.S. monetary base is worth only about two months of its GDP. There has always been a large difference, but it’s larger than ever now. This situation is likely fueled by the oversized banking sector in Switzerland as well as the refuge currency role of the Swiss franc. The latter is particularly true in times of uncertainty, including the uncertainty of its neighbors’ currency, the euro.

How this graph was created: Search for and select “Swiss monetary base” and click “Add to Graph.” From the “Edit Graph” panel, add a series by searching for “Switzerland GDP,” taking the quarterly series with nominal data, and applying formula a/b. Then, from the “Add Line” tab, search for and select “monetary base,” add a series by searching for “GDP” again taking the nominal series and applying formula a/b/1000. Finally, adjust the sample period to start in 1980.

Suggested by Christian Zimmermann.

View on FRED, series used in this post: BOGMBASE, CPMNACSAB1GQCH, GDP, SNBMONTBASE

What can we claim about initial claims?

Keeping track of initial unemployment insurance claims

Initial unemployment claims is a much-watched indicator of the economy. It counts how many people have become eligible for unemployment insurance compensation in a particular week. The data are available quickly and at a high frequency (weekly), but the series has the disadvantage of being highly volatile. This is why FRED also offers a four-week moving average, shown in the graph above: Simply, it’s the average of the past four weeks. Included in the graph is also a red line that indicates the lowest value of this statistic in the course of its history—in May 1969. Currently, claims are around 230,000 per week; and, while this is low, it was lower for 126 weeks early in the sample period.

Of course, the population was much smaller in the 1960s, so the current statistics are even more impressive than they first appeared. Which is what the second graph shows, after dividing new claims by population. The red line indicates the lowest point before recent years, which occurred in April 2000. That low point has clearly been beaten—ever since May 2015, in fact. Keep in mind, though, this statistic is only part of the labor market picture. For example, average unemployment duration is still elevated (see a previous blog post on this). Also, unemployment insurance eligibility requirements may vary over time and, thus, distort the statistic.

How these graphs were created: Search for and select the 4-week moving average for “initial claims” and click “Add to Graph.” From the “Edit Graph” panel, use the “Add Line” feature to create a “user-defined line” and enter 179,000 for the start and end values. For the second graph, edit the first graph by adding a series to the first line, searching for “civilian population” and applying the formula a/(b*1000). Use the “Add Line” feature to create a “user-defined line,” and enter 0.00223 for the start and end values.

Suggested by George Fortier.

View on FRED, series used in this post: IC4WSA, LNU00000060


Back to Top