Federal Reserve Economic Data

The FRED® Blog

Crowds in the air

Graphing airfares and passenger load factors

Do you feel lucky if no one sits beside you on an airplane? Lucky might be the right word for it: According to data provided by the U.S. Bureau of Transportation Statistics (BTS), the chance of having an empty seat next to you has been getting slimmer over time.

Over the past two decades, passenger load factors in the U.S. have been rising as air travel has gotten more crowded. Roughly speaking, passenger load factor is the average percentage of airplane seats occupied across all flights. A load factor of 0 percent indicates an empty flight while a load factor of 100 percent indicates a full flight. Because each flight has different capacities and travel distances, the BTS adjusts the official load factor statistics by the size of the airplanes as well as the flight distances.

The top graph plots the load factor for U.S. domestic flights. The red line and blue line indicate the load factors before and after seasonal adjustment, respectively. The load factor increased gradually from about 70 percent in 2000 to about 85 percent in 2017. This trend implies that the airline industry has taken better advantage of the capacity of airplanes over time. In addition, the significant gaps between the lines reflect the seasonality of air travel, with summers being the most popular time to fly. However, these seasonal gaps seem to shrink over time. One possible explanation could be that the airline industry has improved its use of airplane capacity in the winter when traveling is less popular.

This 15-percentage-point increase in load factor, from 70 percent to 85 percent, likely makes a passenger’s flight feel significantly more crowded. Consider a single-aisle aircraft with three seats on each side of the aisle. If the load factor were 66 percent, which is not far from the 70 percent number in 2000, then ideally no one would have to sit in the middle seat and there would be an empty seat between each passenger. As the load factor increases from 66 percent to 85 percent, the percentage of passengers sitting next to an empty seat drops from 100 percent to 30 percent. This more-confined experience could make some passengers more willing to pay extra fees for preferred seats or seats with better legroom.

However, the higher load factors produce some benefits for consumers: The average airfare has been lower because airlines have been putting their capacity to better use. The bottom graph shows that airfares have decreased over the past several years.

How these graphs were created: For the top graph, search for “domestic load factor, scheduled passenger flights,” check the seasonally adjusted and not seasonally adjusted series, and click the “Add to Graph” button near the top of the search results. For the bottom graph, search for “consumer price index for all urban consumers: airline fare,” check the seasonally adjusted series, and click the “Add to Graph” button.

Suggested by YiLi Chien.

View on FRED, series used in this post: CUSR0000SETG01, LOADFACTORD, LOADFACTORDD11

Extra, extra, read all about it!

There’s more to the story than headline unemployment

Earlier this month, the Bureau of Labor Statistics reported September’s unemployment rate was 4.2% and the number of unemployed persons decreased by over 300,000. Does that mean every one of those 300,000 individuals found a job? Many Americans view decreases in unemployment positively, but there’s more to the unemployment story than just the headline. The reported unemployment rate is a proportion of members of the civilian labor force who have actively but unsuccessfully searched for a job in the past four weeks. This measure is frequently criticized for ignoring two segments of the population: those who have given up searching for work (i.e., discouraged workers) and those who would accept a job if offered one but aren’t actively searching (i.e., marginally attached workers). Moreover, those working part-time are counted equally in employment data as those working full time, regardless of what they’d prefer.

Because the BLS understands the shortcomings of the headline rate, they record unemployment data for all these groups and release six different measures of unemployment. The graph shows some of these measures for the state of Missouri. The teal line on the bottom shows the headline unemployment rate, and the green line adds the percentage of discouraged workers. Notice that the number of discouraged workers increases after a spike in the headline rate. While headline unemployment reached a peak in the first quarter of 2010, the number of discouraged workers was highest throughout 2011 and 2012, reaching 1.3%. The increase in individuals not actively searching for a job indicates some of the social costs of unemployment: pessimism toward the labor market and a pervasive belief that jobs are increasingly difficult to find, causing workers to stop searching for jobs in the first place. This collective shift may take years to improve, explaining the delayed yet sustained increase.

The measure shown by the dark line includes the percentage of marginally attached workers but does not include discouraged workers. This category includes individuals who have searched for a job in the past 12 months, but not in the past four weeks. This measure of marginally attached workers as a percentage of the whole remains fairly consistent over time. The final segment of the unemployed population, shown by the blue line, adds the percentage of individuals employed part-time for economic reasons: They desire a full-time job, but work fewer than 35 hours per week. We expect the proportion of these individuals to increase during times of economic downturn as employers may cut hours before firing employees. However, the percentage is highest in the second quarter of 2010, at 5.4%, and not earlier in the Great Recession. Reasons for this delay may be economic uncertainty and the possibility of workers on contracts that delay employers’ responses. Furthermore, many workers may have been satisfied with a part-time job during the Great Recession, but then began reporting their part-time work as involuntary as the recovery started and the outlook improved.

How this graph was created: Search for “unemployment in Missouri” and select the seasonally adjusted series. From the “Edit Graph” tab, click “Add line” and search for “unemployed plus discouraged Missouri” and select the relevant series. Add a third line and search for “unemployed plus marginally attached Missouri” and select the series. Finally, add a fourth line and search for “unemployed plus part-time Missouri.” Change the start date to 01-01-2004.

Suggested by Maria Hyrc and Christian Zimmermann.

View on FRED, series used in this post: MOUR, U4UNEM4MO, U5UNEM5MO, U6UNEM6MO

Three views of the U.S. trade deficit

Minding units and considering services

Consider the graph above, which shows the U.S. trade balance. It looks like things are seriously heading south, with a deficit that’s ten times larger than it was 25 years ago. Is it really that bad? For one thing, the economy as a whole has grown significantly over this period, and prices have increased, too. To address these biases, we should divide the trade balance by our favorite nominal index, nominal GDP. The result is the graph below.

Now that the units are percentages of GDP, we can see that the deficit is five times as large as it was 25 years ago, not ten times. And it has actually improved since the previous recession, to a little more than three times its size, topping out at –4% of GDP. But wait, there’s more: International trade doesn’t pertain to goods alone; it also involves services. And here, the United States actually enjoys a surplus. So, if you redo the second graph with the trade balance for goods and services, you obtain the graph below:

Finally, we see that the current trade deficit is at about 3% of GDP. Is that a lot? Actually, a deficit isn’t necessarily bad. See a previous blog post on the topic.

How these graphs were created: For the first graph, simply search for “trade balance” and take the series that pertains only to goods. For the second graph, use the first and then go to the “Edit Graph” panel: From there, add “nominal GDP” and apply the formula a/b/10*12. (The idea is to divide by 1,000 to put both series into the same units and then multiply by 100 to obtain results in percentages, which reduces to simply dividing by 10. Multiplying by 12 changes the trade balance’s monthly frequency to an annual frequency, to match nominal GDP’s annual frequency.) For the the third graph, replace the trade balance for goods with the trade balance for goods and services.

Suggested by Christian Zimmermann.

View on FRED, series used in this post: BOPGSTB, BOPGTB, GDP


Back to Top