The Boom Forecasted

by: Econbrowser

Editor's note: This article was originally published on 12/23/2014 by Menzie Chinn here.

A year ago, Jeff Frieden and I forecast (along with others) in Politico Magazine a boom in output. In light of the Dec. 23's third release for 2014Q3 real GDP, clocking in at 5% SAAR [1], it's reasonable to ask, are we there yet?

Figure 1 depicts the upward revision in GDP from advance to final, in the context of the last two years.


Figure 1: GDP, in billions of Ch.2009$, SAAR, 3rd release (bold blue), and advance (red). Macroeconomic Advisers tracking forecast as of 12/23 for 2014Q4 (dark blue square). Vertical dashed line at maximal estimated impact on GDP growth according to Macroeconomic Advisers (2/20/2014). Source: BEA 2014Q3 3rd and advance releases, and Macroeconomic Advisers.

Notice that there is a slowdown in 2013Q2 when Macroeconomic Advisers indicated the maximal impact on growth. (Notice that some observers have argued that since the growth impact was only estimated for two quarters, then the sequester had minimal impact. Of course, those with a familiarity with math recognize that a transient impact on growth has a long-term impact on the level unless the impact on growth is offset.)

Where does a lot of the growth revision come from? The answer is in great part - services.


Figure 2: Services consumption, in billions of Ch.2009$, SAAR, 3rd release (bold blue), and advance (red). Vertical dashed line at maximal estimated impact on GDP growth, according to Macroeconomic Advisers (2/20/2014). Source: BEA 2014Q3 3rd and advance releases.

The estimates of services consumption have been particularly problematic with the advent of ACA implementation (which by the way seems to be taking uninsured rates to new lows). But both health and non-health services consumption were revised upward.

One prediction we made was that investment would surge, given corporate profit levels and the decrease in policy uncertainty. This has come to pass only in part.


Figure 3: Log nonresidential fixed investment to GDP, both in billions of Ch.2009$, SAAR, 3rd release, rescaled to 1986Q1=0 (blue, left scale), and policy uncertainty index (pink, right scale). Policy uncertainty for 2014Q4 is through 12/23. Source: BEA 2014Q3 3rd release, and Baker, Bloom and Davis.

Notice that the investment series is business fixed investment, expressed as a log ratio to GDP. The series only has an interpretation in terms of cumulative differential growth rates since 1986Q1. (For a cautionary tale regarding the use of chain weighted quantities, see this post) By that metric, with the 0.8% upward revision of investment relative to advance, the 2014Q3 differential almost matches the differential of the last two peaks, with an acceleration in the last three quarters (as we predicted). That surge has occurred against a backdrop of reduced policy uncertainty.

That being said, the net ratio is probably lower since economic depreciation of capital has probably accelerated over the past decade as a higher proportion of investment is in shorter lived ICT capital and software.

Interestingly, radio silence on the release from the Joint Economic Committee (Republicans), who typically religiously send me critiques of the latest economic releases. Furman/CEA provides their assessment here.

A final observation: while growth is now proceeding more rapidly, with solid fundamentals in place for growth in 2015, it's important to recall the CBO estimated output gap remains large, at 3.3% (log terms). Moreover, any number of international developments could still derail growth. As could another bout of fiscal brinksmanship (e.g., [2]).