Federal Reserve Economic Data: Your trusted data source since 1991

The FRED® Blog

CO2 in the air: How does it get there?

CO2 emissions by fuel type and sector

In a previous post, we looked at carbon emissions by fuel type broken down by different economic sectors. Today, we slice the data another way: We look at each economic sector and break down their emissions by fuel type. The first graph shows that the big emitters are transportation, electric power generation, and industry. Overall emissions have tended to decline, mostly thanks to a decline from power generation.

The next graph shows the commercial sector. Overall, it emits relatively little CO2 and all fuel types seem to be on the decline. The recent surge in gasoline is most likely due to a reclassification of some sub-sectors into the commercial sector.

The next graph, which shows emissions from the industrial sector, isn’t very enlightening, as the largest fuel type is “Other.” But all fuel types are emitting less, except for distillate fuels such as diesel.

Electric power generation is traditionally the largest emitter, so it’s particularly relevant to consider its fuel composition. A clear majority of its emissions come from coal, but this is now on a steady decline. Natural gas has increased, but overall emissions from this sector have been decreasing.

Our last two graphs consider the transportation and residential sectors: Clearly, the transportation sector is very heavily into petroleum, with a slight upward trend in its emissions. The residential sector is heavily into natural gas, plus a bit of petroleum, with a slight downward trend.

How these graphs were created: For the first, search for “carbon dioxide emissions all fuels,” use the side bar to restrict results to “nation,” select the series shown here, and click “Add to Graph.” From the “Edit Graph” panel, use the “Format” tab to select graph type “Area” and stacking “Normal.” The five other graphs are built similarly by searching for “carbon dioxide emissions” and the respective sector, including only series where the units are million metric tons. Note: The “Format” tab also allows you to choose colors for the series, which is useful for making the colors for the fuels consistent across graphs.

Suggested by Christian Zimmermann.

View on FRED, series used in this post: EMISSCO2TOTVCCTOUSA, EMISSCO2TOTVECCOA, EMISSCO2TOTVECNGA, EMISSCO2TOTVECPEA, EMISSCO2TOTVECTOUSA, EMISSCO2TOTVICTOUSA, EMISSCO2TOTVRCCOA, EMISSCO2TOTVRCNGA, EMISSCO2TOTVRCPEA, EMISSCO2TOTVRCTOUSA, EMISSCO2TOTVTCCOA, EMISSCO2TOTVTCNGA, EMISSCO2TOTVTCPEA, EMISSCO2TOTVTCTOUSA, EMISSCO2VCLCCBA, EMISSCO2VCLICBA, EMISSCO2VDFCCBA, EMISSCO2VDFICBA, EMISSCO2VKSCCBA, EMISSCO2VLUICBA, EMISSCO2VMGCCBA, EMISSCO2VMGICBA, EMISSCO2VRFCCBA, EMISSCO2VRFICBA

Have you heard the news? News can affect markets

The effects of economic news on expectations of future financial performance

FRED’s all about data, which economists often use to conduct or test their research. So let’s look at some of that research…

In a recent St. Louis Fed working paper, economists Maximiliano Dvorkin, Juan M. Sanchez, Horacio Sapriza, and Emircan Yurdagul study how the arrival of news affects emerging markets. They use a logic from a 2006 paper by Beaudry and Portier to identify news events—aka “shocks.” The idea is to compare a financial index that captures the expected future performance of the economy with a measure of current performance. They identify “good news” when the expected performance variable improves without any proportional improvement in the current performance variable. On the flip side, they identify “bad news” when the expected performance variable declines without any proportional decline in the current performance variable.

Because their research focuses on emerging markets, they use the JPMorgan Emerging Market Bond Index (EMBI) spread, which captures the risk of sovereign default, as their measure of future performance. They show a connection between the arrival of bad news and an increase in the EMBI spread that can’t be accounted for by current data. They also find that these shocks are important in accounting for fluctuations in these emerging economies and that these economies can’t shield themselves from news shocks by extending the maturity of their debt.

Now, back to FRED: Data can be used to test and illustrate the logic behind this research. The graph shows (in blue) the St. Louis Fed Economic News Index that’s used to predict the value of current real GDP before the BEA releases the official data. Assuming this index is good at capturing current news, we should see a strong correlation between this index and a financial index affected by the future performance of the economy. The index we chose (in red) is the S&P 500: This measure of the value of the stock market, as a measure of the expected performance of U.S. companies, serves as our measure of future performance.

The graph shows that the S&P 500 and Economic News Index move closely together, which suggests the logic is correct and that additional research could identify how news affects the U.S. economy.

How this graph was created: Search for and select “St. Louis Fed Economic News Index: Real GDP Nowcast.” From the “Edit Graph” panel, use the “Add Line” tab to search for and select the S&P 500 series; then click “Add data series.” From there (the “Edit Lines” tab), adjust the units to “Percent Change from Year Ago” for comparability with the news index. Now, both lines will be on the same graph, but their scales are quite different. To better compare the two, you can select “Format” and change the y-axis position to “Right” for the S&P 500 line.

Suggested by Ryan Mather and Juan Sánchez.

View on FRED, series used in this post: SP500, STLENI

Do government dollars drive recovery?

The conventional wisdom and data behind government spending during recessions

Conventional wisdom suggests that, once you determine the appropriate level of government spending on goods and services, this level should grow more or less in line with the growth of the broader economy. Keeping the growth rate of government spending stable over the business cycle helps stabilize the business cycle.

But let’s see what the data show: The FRED graph above plots (1) the percentage change year-to-year of total government spending on goods and services and (2) the employment-to-population ratio. The three shaded regions in the graph represent periods of recession, each characterized by a rapid decline in employment followed by a gradual recovery. But the growth in nominal government spending wasn’t the same across these three recessions: It decelerates in the early 1990s recession, remains relatively stable in the early 2000s recession, and declines precipitously into negative territory in the most recent recession.

The recessions themselves were also different: The recession in the early 2000s was noticeably mild. Was this in part due to the stable pace of government spending? In contrast, the 2007-09 recession was very deep and had a very slow recovery. Was this in part due to the unprecedented cuts in government spending? At the state and local level, these cuts were made largely in response to diminished state tax revenue and the inability to issue debt. At the federal level, they were motivated more by unwillingness to expand the federal debt even further.

Austerity during a downturn may have its merits. But the fiscal retrenchment during the most recent recession almost surely contributed to the recession’s severity and very slow recovery. Given that interest rates and inflation remained unusually low, it seems difficult to justify the sharp cyclical cuts in government spending that took place at that time. If a smaller government spending program is a long-term policy goal, the textbook recommendation is that this policy should be implemented only after the economy has fully recovered from recession.

How this graph was created: Search for “Government Consumption Expenditures and Gross Investment (GCE).” From the “Edit Graph” panel, change “Units” to “Percent Change from Year Ago.” Select “Add Line” in the same editing panel and search for the “Employment Population Ratio: 25 – 54 years.” Select the first result. Under “Edit Line 2” (still in the same editing panel), change “Units” to “Percent.” Finally, change the start and end dates on the graph to “1990-01-01” and “2019-12-01.”

Suggested by David Andolfatto and Mahdi Ebsim.

View on FRED, series used in this post: GCE, LNS12300060


Subscribe to the FRED newsletter


Follow us

Back to Top