This GDP report is whispering some really important messages to us

October 27th, 2017 at 10:45 am

A bunch of notable points from this morning’s advanced GDP report for 2017Q3. The economy’s growing a good clip (though not at the 3% headline number), with no price pressures in sight. Consumer spending is solid, boosted by job and earnings growth as well as, I suspect, a wealth effect from the stock market (stock market gains are not in GDP, but if people feel wealthier, they spend a little bit more). The hurricanes didn’t faze the data flow (note: Puerto Rico is not counted in national GDP), and a few noisy factors, especially an inventory bump, helped boost the number (I smooth out the noise below).

Yes, the White House is going to brag on this report, as pretty much any White House would. But what they should see here is yet another great reason to not screw around with the Fed chair. Leave Yellen where she is, Trump, and go tweet about your awesome golf skills, or something.

Also, why, why, why—if the economy is percolating along at a solid, steady clip, do we need big-ass, wasteful tax cut? That’s a rhetorical question, but other than “because we won and that’s what our donors want!,” I’ve not heard anything like a decent answer to that question.

Look at the yearly, not annualized quarterly rates

First, readers know I’m all about discounting the annualized quarterly growth rates, which as you see below are choppy. So, while the White House will crow about the topline 3% number (the real, annualized growth from q2 to q3), I emphasize the 2.3% year-over-year growth rate. The figure shows both measures and the important bit is the way the yearly rate smooths out the bips/bops of the annualized quarterly one.

Source: BEA

Punchline: the underlying, trend growth rate of real GDP is around 2%, though slightly north of that, and the last six quarters do show a nice, steady acceleration of the yearly growth rate (I’ll get back to the price column of the table below):

Source: BEA

We’re realizing our diminished potential

The Congressional Budget Office estimates potential GDP, i.e., what the level of the economy’s output should be at full utilization of our stock of labor and capital, given our productivity (the latter being how efficiently were transforming said inputs into outputs). As of 2017q3, nominal GDP is back to potential, so by this measure, we’re at full employment (see figure; this is the first quarter in the expansion that both real and nominal GDP beat CBO’s potential). You’d get the same result from looking at the 4.2% unemployment rate and various estimates of the lowest rate believed to be consistent with stable prices.

Sources: BEA, CBO

That “believed to be” is important. As I’ll get to in a moment, actual inflation is inconsistent with these conclusions. If things were all that tight, we should be seeing faster price growth, which we’re not…at least not yet.

But there’s a more profound point in play here. The last bar in the figure is CBO’s 2007 forecast of potential GDP for this last quarter, and it’s about $2 trillion above its most recent potential estimate. That’s decline represents an income loss of over $6,000 per person.

What the heck happened? I’m working on a longer piece that tries to answer that question, but some economists, including myself, view part of that loss to stem from “hysteresis,” the lasting, structural damage done to the economy’s inputs due to long periods of persistent weakness. Think of somebody sitting out the recovery because there’s not enough opportunity for work (labor demand is too weak), and their skills, attitude, etc. atrophies such that they’re eventually no longer part of the labor force (here’s an NYT piece w/some evidence).

Phlat Phil

The Phillips Curve is the correlation between unemployment or potential GDP and inflation and man, is that correlation low. As you see in the second column of that little table above, core inflation is decelerating as real GDP accelerates.

Barrels of digital ink have been spilled trying to understand why Phil’s so phlat but let me save you weeks of your life by cutting to the chase: nobody knows.

I can tell that you that the curve has been flat for a while, at least at the national level (you can find some correlations in cities), such that claims that it’s temporarily out of commission due to some cherry-picked factor don’t seem very convincing. If we were to shave with Occam’s razor, we’d have to recognize that the simplest explanation is that we may well not be at full employment, and CBO’s potential estimate is too low, for which there’s actually some compelling evidence (all to be explored in my forthcoming paper).

It’s just one, advanced (meaning it will be revised) GDP report, but as a veteran GDP-report whisperer, let me tell you what this one is saying: reappoint Yellen and Fed, wherein she must continue to apply great patience in normalizing rates.

Print Friendly, PDF & Email

9 comments in reply to "This GDP report is whispering some really important messages to us"

  1. Smith says:

    Enough already with the Phillips curve. Fed bankers use it only because they are intellectually bankrupt, and there is no one to bail them out.
    Inflation is caused by higher prices, not wages (which contribute on average 18 percent of product price or cost of services across all industries). Sellers have the greater control over prices than buyers.
    Inflation results from lack of competition, speculation, cartels, natural monopolies, insurance paid services and products, fraud, natural disasters, world interest and exchange rates, bidding up of prices by inequality wealth laden rich, a gentrification of the economy.
    Wages are determined more by convention than supply and demand. Employers control the means of production, firms with greater than 500 employees control the labor market for over 50% of workers. Minimum wages, competition from immigrants with lower expectations due to undeveloped state of home country’s economy, lower wage demands of women who along with immigrants face discrimination too.
    Rich people, management, and those with more education have greater control over their own wages and so they favor themselves and people more similar to themselves.

    Sometimes supply and demand influences wages, such as influx of females into job market or lower wage immigrants, or demand for tech skills in pre 2000 tech boom. Generally there is a greater influence downward because business addresses wage demands avoiding them.
    Likewise prices go up but rarely down. Business adds new features, so my computer still cost $500 to $1,000 and not $50 to $100. Big question how this affects inflation and productivity numbers. Oil prices went down for sure, but when they go up, for some reason people look at inflation and say slow the economy, increase unemployment.

  2. Bryce says:

    A Philips curve should be based on the prime employment rate:

    When we restrict the time period for which we have data from the J2J data (the second quarter of 2000 until the fourth quarter of 2015), the prime employment rate still explains the most variation. For nominal wages, the prime employment rate explains 85 percent, the unemployment rate explains 65 percent, and the job-switching rate explains 67 percent. For inflation-adjusted compensation, the prime employment rate explains 50 percent, the unemployment rate explains 31 percent, and job switching explains 37 percent.

    • Smith says:

      Great, I hit that point about prim age employment Oct 10 and in many earlier comments too. Not sure why your link says six months and my calculations show 2 years to go. It could be because it sees 6 months as the time to get wages at a minimal 3 % nominal growth, which is still probably 1 percent or less real growth. But the two year estimate is still valid when compared to recent and historical data.
      an excerpt of which I leave here:
      “My calculations show slack in the labor market of even 1 percent less prime age (120 million) workforce participation leaves 1,200,000 waiting to reenter job market. The participation rate is historically low, you have to go back to 1987 to find lower rates for any period of time. It also means 150,000 new job growth trending now leaves two years to go for anything like full employment (100,000 needed for normal population growth).”

  3. jorgensen says:

    The 2007 GDP number was not “real” and no one should be extrapolating from it. It included too much mispricing and outright fraud in housing and finance to be a reliable benchmark.

    In a year when Thaler has won the Nobel, economists should recognize that clinging to 2007 as indicating where the economy should be is just an example of the endowment effect.


    • Smith says:

      It is a conservative argument that the growth lines in everything are fine now because the 2000s were a bubble economy. Summers also uses the bubbleness as an argument for secular stagnation. The opposite is true. Wages, even for most of high income and college educated, stagnated. The 2001 recession was followed by a jobless recovery, and manufacturing lost 3 million jobs in just 3 years. Percent of imports over GDP reached record 6 percent. Bubbly activity may have inflated some numbers, but that only hid the hideous performance of the rest of the economy, so that a normal trend line would have left us where we left off in 2007 anyway. Yet we never bounced back, so there is a $1 trillion missing from the economy every year. That also translates into more than $300 billion missing government revenue.

    • Peter K. says:

      Yes the 2007 number was real, there was no accelerating inflation. There are papers saying the output gap could be as large as 10 percent. Bubbles are different than inflation.

      Mainstream economics is pretty bad when it comes to things which should be in its wheel house: inflation, the output gap, potential GDP, etc.

  4. Pinkybum says:

    “Barrels of digital ink have been spilled trying to understand why Phil’s so phlat but let me save you weeks of your life by cutting to the chase: nobody knows.”

    It’s because we are not at full employment and it is very obvious. The participation rate is still down 2.5% from Jan 1999 for the 25-54 yr old cohort. The employment to Population Ratio is down 3.5% from May 2000 for the 25-54 yr old cohort. There is still at a lot of slack in the employment market.

  5. Peter K. says:

    Stepanie Kelton in the NYTimes:

    McCulley: I think it should be a collaborative venture between the Fed and Congress. Yes, I used the word “collaborative,” which I think applies in a more general way to the relationship between Congress and the Fed.

    The Fed’s operational independence is grounded in the thesis that the legislature cannot be trusted with monetary policy, as the electoral process is inherently biased to inflation, of overstimulating the economy with too much spending relative to taxation, running inflationary budget deficits.

    That simply has not been the case for a long, long time. Yes, we’ve had large deficits, but inflation has been too low, not too high. Thus, I’m not convinced by the argument that strict Fed independence is always and everywhere needed to discipline the fiscal authorities’ inflationary bias.

    Kelton: What about policy today?

    McCulley: If President Trump wants to try to boost real growth from 2 percent to 3 percent, there is no reason that the Fed should actively push back. That doesn’t mean that the Fed shouldn’t or wouldn’t respond if such an acceleration in growth were to finally drive unemployment low enough to generate a loud wage and inflationary response.

    My point is that there is no reason for the Fed to prevent the “experiment,” if Mr. Trump and the Republican Congress want to run it.

    I usually agree with you but it is disingenuous to argue that the Fed will limit potential GDP when discussing macro plans like Gerald Friedman’s evaluation of Sanders’ plan, or the Republican plan to boost growth. If the Fed is going to limit growth then there is a problem with the Fed.

  6. lisa says:

    Even without the CPI and PCE for time periods 1930 to almost 1950, What conclusions can you reach from this chart through 2017 - Especially about PPI and GDP?