Skip to main content

The FRED® Blog

Which states are most invested in trade with China and Canada? The geographic distribution of U.S. exports

If you follow this blog, chances are you’ve run across at least some standard economic theories. For example, (1) countries export what they can produce at a comparative advantage and import the other stuff and (2), with nearly unequivocal agreement, free trade is seen as beneficial overall for trading partners. You may also be following the escalating tensions between the U.S. and its trading partners (China, Canada, Mexico, Europe, etc.) over tariffs enacted by the U.S. to protect import-competing industries and the retaliatory tariffs enacted by the other countries. So let’s see if FRED data can connect a little theory with current events.

To keep it simple, we look at U.S. state exports to Canada and then to China. The first map makes it clear that northern states export more to Canada than other states. This aligns well with standard economic models: The factors that determine trade relationships include distance between countries, incomes of trading countries, common languages, and common borders. But we also see that larger states such as California and Texas are major exporters to Canada, too. In 2016, U.S. states exported goods worth around $193.7 billion to Canada. Michigan and Ohio were the largest exporters, with a combined 20.19% of total state exports to Canada.

The second map shows that many of the major exporters to Canada are also major exporters to China, including California, Texas, Michigan, and Ohio. Several of these states serve as major ports (California, Texas, and Ohio, for instance), which is one potential explanation for why these states are major exporters in both cases. U.S states exported goods worth $93.9 billion to China, with 25.75% of them originating in California and Ohio.

You might be asking yourself why state-level trade patterns matter. One reason is that states aren’t all invested in international trade to the same degree. Trade affects states differently, according to their specific industries and those industries’ exposure to foreign purchases of their products. And these trade patterns can provide insights into how tariffs and other changes in international trade could affect specific U.S. cities, states, and regions.

How these maps were created: Go to GeoFRED and click on “Build New Map.” Under “Tools,” set Region Type to “State” and Data to “Value of Exports to Canada.” Choose data from the year 2016. Repeat the same process for the second map, with the Data option set to “Value of Exports to China.”

Suggested by Asha Bharadwaj and Maximiliano Dvorkin.

Do imports subtract from GDP? A basic explanation of GDP = C + I + G + (X-M)

The typical textbook treatment of GDP is the expenditure approach, where spending is categorized into the following buckets: personal consumption expenditures (C); gross private investment (I); government purchases (G); and net exports (X – M), composed of exports (X) and imports (M). Textbooks often capture this in one relatively simple equation:

GDP = C + I + G + (X – M).

Notice that, here, imports (M) are subtracted. On the surface, this implies that an extra dollar of spending on imports (M) will decrease GDP by one dollar. For example, let’s assume you spend $30,000 on an imported car; because imports are subtracted (e.g., “– M”), the equation seems to imply that $30,000 should be subtracted from GDP. However, this cannot be correct because GDP measures domestic production, so imports (foreign production) should have no impact on GDP.

When the Bureau of Economic Analysis (BEA; see its primer on this topic) measures economic output, it categorizes spending with the National Income and Product Accounts (NIPA). Some of this spending (which is counted as C, I, and G) is spent on imported goods. As such, the value of imports must be subtracted to ensure that only spending on domestic goods is measured in GDP. For example, $30,000 spent on an imported car is counted as a personal consumption expenditure (C), but then the $30,000 is subtracted as an import (M) to ensure that only the value of domestic production is counted. As such, the imports variable (M) functions as an accounting variable rather than an expenditure variable. To be clear, the purchase of domestic goods and services increases GDP because it increases domestic production, but the purchase of imported goods and services has no direct impact on GDP.

The misconception that imports reduce GDP also seems to be implied when the GDP components are stacked using the FRED release view. Notice that the green “net exports” area is negative. This occurs because the dollar value of imported goods and services exceeds the value of exported goods and services. While this aspect of net exports (X – M) is useful for evaluating how international trade affects economic activity, it can be misleading. Like the misleading aspects of the expenditure equation, it suggests (visually) that imports reduce overall GDP. While the graph is not incorrect, it is important to keep in mind that, when calculating GDP, the value of imports is actually subtracted from the other components of GDP (personal consumption expenditures, gross private domestic investment, government consumption expenditures, and gross investment), not exports. Again, it’s important to emphasize that the imports variable (M) is an accounting variable rather than an expenditure variable.

To learn how to create your own GDP stacking graph, see this FRED blog post. For a more complete description of GDP and the expenditures equation, read the September 2018 issue of Page One Economics.

Suggested by Scott Wolla.

View on FRED, series used in this post: GCE, GPDI, NETEXP, PCEC

Is the housing price-rent ratio a leading indicator?

Economic forecasters are always on the lookout for variables that can help predict upcoming recessions. One such variable that has gotten some recent attention is the housing price-rent ratio. As this ratio becomes higher, the rental option becomes more attractive. If it rises high enough, some households might switch from owning their homes to renting them; then the demand for owner-occupied housing would fall. The result is a contraction in the housing market that can have adverse effects on the entire economy. This narrative seems to match well with the behavior of the housing price-rent ratio leading up to the Great Recession. So if the housing price-rent ratio is on the rise again, does that mean it’s cause for concern? Let’s try to evaluate whether the housing price-rent ratio is a reliable leading indicator by graphing it, with data going back to 1975.

To be considered a leading indicator, a variable must change in sign prior to the beginning of each recession. (Recessions, as defined by NBER, are shown by gray shading.) The Great Recession started in December 2007. As we can see, the housing price-rent ratio reached its peak in April 2006, approximately two quarters prior to the start of the recession. In other words, the housing price-rent ratio seems in this case to have been a leading indicator. But for a complete evaluation, all the recession episodes must be examined. In January 1980, the U.S. economy suffered from double-digit inflation. To solve that problem, Paul Volker essentially created a recession. This recession began in January 1980. The housing price-rent ratio peaked in the second quarter of 1979 and then declined. It could again be argued that the price-rent ratio predicted this recession. In July 1981, another recession started. For this recession, whether the housing price-rent ratio correctly indicated a coming recession is less clear. The housing price-rent ratio didn’t suggest an upcoming recession in March 2001, as the ratio steadily increased.

A second condition for a variable to be a leading indicator is that it doesn’t suffer from the false-positive problem. This problem would occur when the house price-rent ratio decreases but no recession occurs. There are a number of instances when the housing price-rent ratio does suffer from this problem.

So it’s not clear whether the housing price-rent ratio qualifies as a leading indicator: It fails to identify some recessions and gives false-positive readings at other times. But in the two major recessions since 1975 (the 1980 and 2007 recessions), the housing market played a leading role; so, these recessions were predicted correctly by the housing price-rent ratio.

How this graph was created: Search for and select the series called “All-Transactions House Price Index for the United States.” Then, in the customize data option of the “Edit Graph” menu, search for and select the series called “Consumer Price Index for All Urban Consumers: Rent of primary residence.” Finally, in the formula tab, enter a/b to divide the home price index by the rent price index.

Suggested by Ryan Mather and Don Schlagenhauf.

View on FRED, series used in this post: CUUR0000SEHA, USSTHPI

Which wages are really increasing? The evolution of wages by sector

Wages are in the news, so we take the opportunity to see how they’ve evolved recently. In the graph above, wages are separated into three large categories: the goods-producing sector, the service-providing sector, and government. Two notions are clear: Wages are generally trending upward, which should come as no surprise because they haven’t been adjusted for inflation. And the most growth is in the service sector and the least is in government.

The graph below adjusts for inflation using the consumer price index (CPI) and looks at the year-to-year change for all three series. This inflation adjustment makes it clear that wages are not always increasing in real terms. In fact, service sector real wages increase more frequently than government real wages, which is how the gap in the first graph can be explained. Of course, this analysis is at a very high level; our Employment Cost Index release tables offer much more detail.

How these graphs were created: From the Employment Cost Index release tables, select Table 2, then check the series you want, and click “Add to Graph.” For the second graph, click on “Edit Graph” and do the following for each line: add series “CPI,” apply formula a/b, and select units “Percent change from year ago.”

Suggested by Christian Zimmermann.

View on FRED, series used in this post: CIS202G000000000I, CIS202S000000000I, CPIAUCSL, ECIGVTWAG


Subscribe to the FRED newsletter


Follow us

Twitter logo Google Plus logo Facebook logo YouTube logo LinkedIn logo
Back to Top