Visualizing GDP: A Small Improvement Over The Second Estimate

by: Doug Short

Note from dshort: The charts in this commentary have been updated to include the Q4 2012 Third Estimate.

The chart below is my way to visualize real GDP change since 2007. I've used a stacked column chart to segment the four major components of GDP with a dashed line overlay to show the sum of the four, which is real GDP itself. That noticeable change in today's 0.4 percent Third Estimate from the 0.1 percent in the Second Estimate was the positive revision to Gross Private Domestic Investment and continued improvement in Net Exports of Goods and Services. The biggest component of GDP, however, Personal Consumption Expenditures, was revised down from the Second Estimate.

My data source for this chart is the Excel file accompanying the BEA's latest GDP news release (see the links in the right column). Specifically, I used Table 2: Contributions to Percent Change in Real Gross Domestic Product.

Over the time frame of this chart, the Personal Consumption Expenditures (PCE) component has shown the most consistent correlation with real GDP itself. When PCE has been positive, GDP has been positive, and vice versa. In the latest GDP data, the contribution of PCE came at 1.28 of the 0.4 real GDP (down from 1.47 in the Second Estimate).

Note: The conventional practice is to round GDP to one decimal place, the latest at 0.4. The 0.38 GDP in the chart above is the real GDP calculated to two decimal places based on the BEA chained 2005 dollar data series.

Here is a close look at the contribution changes from the Advance to Third Estimates. The upward revisions to Private Investment and Net Exports stand out. However, as I pointed out earlier, Personal Consumption Expenditures, by far the biggest percent of GDP, has been revised downward. Note that the drop in PCE is primarily in Services. That is not good news for employment, since service jobs are a substantial source of employment.

Here a similar look at the contribution changes across the four quarters of 2012.

As for the negative 1.52 contribution from private inventories, let's look at this component over the last twelve quarters. It is highly volatile and, in my view, not very useful in forecasting GDP trends:

2.23, 0.07, 1.97, -1.61, -0.54, 0.01, -1.07, 2.53, -0.39, -0.46, 0.73, -1.52

The decline in defense spending, no doubt a disturbing development for militant conservatives, was not surprising given the war drawdown. This trend may also be in part the result of prior planning for sequestration, which has now begun.

As for the role of Personal Consumption Expenditures (PCE) in GDP and how it has increased over time, here is a snapshot of the PCE-to-GDP ratio since the inception of quarterly GDP in 1947. The Q4 2012 ratio is 70.72%, fractionally off the all-time high of 71.15% in Q1 2011, but fractionally higher than last quarter's 70.46%. From a theoretical perspective, there is a point at which personal consumption as a percent of GDP can't really go any higher. At the low 70 percents, we may be approaching that level.

The one caution I have regarding today's 1.28% contribution of PCE to GDP (a slight downward revision from the 1.47 in the Second Estimate) is the likelihood that tax planning strategies to avoid the expected 2013 tax hike were a key factor in the 2012 end-of-year spending increase. Q1 2013 will probably show the opposite effect, a phenomenon I've highlighted before in my graphs of personal income.

Let's close with a look at the inverse behavior of PCE and Gross Private Investment (GPDI) during recessions (note my use of different vertical scales to facilitate the overlay). PCE generally increases as a percent of GDP whereas Private Investment declines. That is not what we're seeing in the current data. I've plotted the two with different vertical axes (PCE on left, GPDI on the right) to highlight the frequent inverse correlation.

I'll update these charts when the Q1 2013 Advance Estimate is released next month.